Bump version: 0.25.1-beta.4 → 0.25.1

### Bug Fix: Undefined Values in Nullable Fields

**Issue**: When inserting data with `undefined` values into nullable
fields, LanceDB was incorrectly coercing them to default values (`false`
for booleans, `NaN` for numbers, `""` for strings) instead of `null`.

**Fix**: Modified the `makeVector()` function in `arrow.ts` to properly
convert `undefined` values to `null` for nullable fields before passing
data to Apache Arrow.

fixes: #2645

**Result**: Now `{ text: undefined, number: undefined, bool: undefined
}` correctly becomes `{ text: null, number: null, bool: null }` when
fields are marked as nullable in the schema.

**Files Changed**: 
- `nodejs/lancedb/arrow.ts` (core fix)
- `nodejs/__test__/arrow.test.ts` (test coverage)

- This ensures proper null handling for nullable fields as expected by
users.

---------

Co-authored-by: Will Jones <willjones127@gmail.com>

.bumpversion.toml

@@ -1,5 +1,5 @@
 [tool.bumpversion]
-current_version = "0.21.2"
+current_version = "0.22.1-beta.3"
 parse = """(?x)
     (?P<major>0|[1-9]\\d*)\\.
     (?P<minor>0|[1-9]\\d*)\\.

.github/workflows/docs.yml

@@ -56,22 +56,11 @@ jobs:
         with:
           node-version: 20
           cache: 'npm'
-          cache-dependency-path: node/package-lock.json
       - name: Install node dependencies
         working-directory: node
         run: |
           sudo apt update
           sudo apt install -y protobuf-compiler libssl-dev
-      - name: Build node
-        working-directory: node
-        run: |
-          npm ci
-          npm run build
-          npm run tsc
-      - name: Create markdown files
-        working-directory: node
-        run: |
-          npx typedoc --plugin typedoc-plugin-markdown --out ../docs/src/javascript src/index.ts
       - name: Build docs
         working-directory: docs
         run: |

.github/workflows/docs_test.yml

@@ -24,7 +24,8 @@ env:
 jobs:
   test-python:
     name: Test doc python code
-    runs-on: ubuntu-24.04
+    runs-on: warp-ubuntu-2204-x64-8x
+    timeout-minutes: 60
     steps:
     - name: Checkout
       uses: actions/checkout@v4
@@ -58,51 +59,3 @@ jobs:
       run: |
         cd docs/test/python
         for d in *; do cd "$d"; echo "$d".py; python "$d".py; cd ..; done
-  test-node:
-    name: Test doc nodejs code
-    runs-on: ubuntu-24.04
-    timeout-minutes: 60
-    strategy:
-      fail-fast: false
-    steps:
-    - name: Checkout
-      uses: actions/checkout@v4
-      with:
-        fetch-depth: 0
-        lfs: true
-    - name: Print CPU capabilities
-      run: cat /proc/cpuinfo
-    - name: Set up Node
-      uses: actions/setup-node@v4
-      with:
-        node-version: 20
-    - name: Install protobuf
-      run: |
-        sudo apt update
-        sudo apt install -y protobuf-compiler
-    - name: Install dependecies needed for ubuntu
-      run: |
-        sudo apt install -y libssl-dev
-        rustup update && rustup default
-    - name: Rust cache
-      uses: swatinem/rust-cache@v2
-    - name: Install node dependencies
-      run: |
-        sudo swapoff -a
-        sudo fallocate -l 8G /swapfile
-        sudo chmod 600 /swapfile
-        sudo mkswap /swapfile
-        sudo swapon /swapfile
-        sudo swapon --show
-        cd node
-        npm ci
-        npm run build-release
-        cd ../docs
-        npm install
-    - name: Test
-      env:
-        LANCEDB_URI: ${{ secrets.LANCEDB_URI }}
-        LANCEDB_DEV_API_KEY: ${{ secrets.LANCEDB_DEV_API_KEY }}
-      run: |
-        cd docs
-        npm t

.github/workflows/nodejs.yml

@@ -79,7 +79,7 @@ jobs:
       with:
         node-version: ${{ matrix.node-version }}
         cache: 'npm'
-        cache-dependency-path: node/package-lock.json
+        cache-dependency-path: nodejs/package-lock.json
     - uses: Swatinem/rust-cache@v2
     - name: Install dependencies
       run: |
@@ -137,7 +137,7 @@ jobs:
       with:
         node-version: 20
         cache: 'npm'
-        cache-dependency-path: node/package-lock.json
+        cache-dependency-path: nodejs/package-lock.json
     - uses: Swatinem/rust-cache@v2
     - name: Install dependencies
       run: |

.github/workflows/pypi-publish.yml

@@ -56,7 +56,7 @@ jobs:
           pypi_token: ${{ secrets.LANCEDB_PYPI_API_TOKEN }}
           fury_token: ${{ secrets.FURY_TOKEN }}
   mac:
-    timeout-minutes: 60
+    timeout-minutes: 90
     runs-on: ${{ matrix.config.runner }}
     strategy:
       matrix:
@@ -64,7 +64,7 @@ jobs:
           - target: x86_64-apple-darwin
             runner: macos-13
           - target: aarch64-apple-darwin
-            runner: macos-14
+            runner: warp-macos-14-arm64-6x
     env:
       MACOSX_DEPLOYMENT_TARGET: 10.15
     steps:

.gitignore

@@ -31,9 +31,6 @@ python/dist
 *.node
 **/node_modules
 **/.DS_Store
-node/dist
-node/examples/**/package-lock.json
-node/examples/**/dist
 nodejs/lancedb/native*
 dist
 

CLAUDE.md

@@ -11,14 +11,70 @@ Project layout:
 * `nodejs`: The Typescript bindings, using napi-rs
 * `java`: The Java bindings
 
-(`rust/ffi` and `node/` are for a deprecated package. You can ignore them.)
-
 Common commands:
 
-* Check for compiler errors: `cargo check --features remote --tests --examples`
-* Run tests: `cargo test --features remote --tests`
-* Run specific test: `cargo test --features remote -p <package_name> --test <test_name>`
-* Lint: `cargo clippy --features remote --tests --examples`
+* Check for compiler errors: `cargo check --quiet --features remote --tests --examples`
+* Run tests: `cargo test --quiet --features remote --tests`
+* Run specific test: `cargo test --quiet --features remote -p <package_name> --test <test_name>`
+* Lint: `cargo clippy --quiet --features remote --tests --examples`
 * Format: `cargo fmt --all`
 
 Before committing changes, run formatting.
+
+## Coding tips
+
+* When writing Rust doctests for things that require a connection or table reference,
+  write them as a function instead of a fully executable test. This allows type checking
+  to run but avoids needing a full test environment. For example:
+    ```rust
+    /// ```
+    /// use lance_index::scalar::FullTextSearchQuery;
+    /// use lancedb::query::{QueryBase, ExecutableQuery};
+    ///
+    /// # use lancedb::Table;
+    /// # async fn query(table: &Table) -> Result<(), Box<dyn std::error::Error>> {
+    /// let results = table.query()
+    ///     .full_text_search(FullTextSearchQuery::new("hello world".into()))
+    ///     .execute()
+    ///     .await?;
+    /// # Ok(())
+    /// # }
+    /// ```
+    ```
+
+## Example plan: adding a new method on Table
+
+Adding a new method involves first adding it to the Rust core, then exposing it
+in the Python and TypeScript bindings. There are both local and remote tables.
+Remote tables are implemented via a HTTP API and require the `remote` cargo
+feature flag to be enabled. Python has both sync and async methods.
+
+Rust core changes:
+
+1. Add method on `Table` struct in `rust/lancedb/src/table.rs` (calls `BaseTable` trait).
+2. Add method to `BaseTable` trait in `rust/lancedb/src/table.rs`.
+3. Implement new trait method on `NativeTable` in `rust/lancedb/src/table.rs`.
+    * Test with unit test in `rust/lancedb/src/table.rs`.
+4. Implement new trait method on `RemoteTable` in `rust/lancedb/src/remote/table.rs`.
+    * Test with unit test in `rust/lancedb/src/remote/table.rs` against mocked endpoint.
+
+Python bindings changes:
+
+1. Add PyO3 method binding in `python/src/table.rs`. Run `make develop` to compile bindings.
+2. Add types for PyO3 method in `python/python/lancedb/_lancedb.pyi`.
+3. Add method to `AsyncTable` class in `python/python/lancedb/table.py`.
+4. Add abstract method to `Table` abstract base class in `python/python/lancedb/table.py`.
+5. Add concrete sync method to `LanceTable` class in `python/python/lancedb/table.py`.
+    * Should use `LOOP.run()` to call the corresponding `AsyncTable` method.
+6. Add concrete sync method to `RemoteTable` class in `python/python/lancedb/remote/table.py`.
+7. Add unit test in `python/tests/test_table.py`.
+
+TypeScript bindings changes:
+
+1. Add napi-rs method binding on `Table` in `nodejs/src/table.rs`.
+2. Run `npm run build` to generate TypeScript definitions.
+3. Add typescript method on abstract class `Table` in `nodejs/src/table.ts`.
+4. Add concrete method on `LocalTable` class in `nodejs/src/native_table.ts`.
+    * Note: despite the name, this class is also used for remote tables.
+5. Add test in `nodejs/__test__/table.test.ts`.
+6. Run `npm run docs` to generate TypeScript documentation.

Cargo.toml

@@ -1,10 +1,5 @@
 [workspace]
-members = [
-    "rust/lancedb",
-    "nodejs",
-    "python",
-    "java/core/lancedb-jni",
-]
+members = ["rust/lancedb", "nodejs", "python", "java/core/lancedb-jni"]
 # Python package needs to be built by maturin.
 exclude = ["python"]
 resolver = "2"
@@ -20,16 +15,14 @@ categories = ["database-implementations"]
 rust-version = "1.78.0"
 
 [workspace.dependencies]
-lance = { "version" = "=0.32.1", "features" = [
-    "dynamodb",
-], "tag" = "v0.32.1-beta.2", "git" = "https://github.com/lancedb/lance.git" }
-lance-io = { "version" = "=0.32.1", "tag" = "v0.32.1-beta.2", "git" = "https://github.com/lancedb/lance.git" }
-lance-index = { "version" = "=0.32.1", "tag" = "v0.32.1-beta.2", "git" = "https://github.com/lancedb/lance.git" }
-lance-linalg = { "version" = "=0.32.1", "tag" = "v0.32.1-beta.2", "git" = "https://github.com/lancedb/lance.git" }
-lance-table = { "version" = "=0.32.1", "tag" = "v0.32.1-beta.2", "git" = "https://github.com/lancedb/lance.git" }
-lance-testing = { "version" = "=0.32.1", "tag" = "v0.32.1-beta.2", "git" = "https://github.com/lancedb/lance.git" }
-lance-datafusion = { "version" = "=0.32.1", "tag" = "v0.32.1-beta.2", "git" = "https://github.com/lancedb/lance.git" }
-lance-encoding = { "version" = "=0.32.1", "tag" = "v0.32.1-beta.2", "git" = "https://github.com/lancedb/lance.git" }
+lance = { "version" = "=0.37.0", default-features = false, "features" = ["dynamodb"] }
+lance-io = { "version" = "=0.37.0", default-features = false }
+lance-index = "=0.37.0"
+lance-linalg = "=0.37.0"
+lance-table = "=0.37.0"
+lance-testing = "=0.37.0"
+lance-datafusion = "=0.37.0"
+lance-encoding = "=0.37.0"
 # Note that this one does not include pyarrow
 arrow = { version = "55.1", optional = false }
 arrow-array = "55.1"
@@ -40,12 +33,12 @@ arrow-schema = "55.1"
 arrow-arith = "55.1"
 arrow-cast = "55.1"
 async-trait = "0"
-datafusion = { version = "48.0", default-features = false }
-datafusion-catalog = "48.0"
-datafusion-common = { version = "48.0", default-features = false }
-datafusion-execution = "48.0"
-datafusion-expr = "48.0"
-datafusion-physical-plan = "48.0"
+datafusion = { version = "49.0", default-features = false }
+datafusion-catalog = "49.0"
+datafusion-common = { version = "49.0", default-features = false }
+datafusion-execution = "49.0"
+datafusion-expr = "49.0"
+datafusion-physical-plan = "49.0"
 env_logger = "0.11"
 half = { "version" = "2.6.0", default-features = false, features = [
     "num-traits",
@@ -62,12 +55,11 @@ rand = "0.9"
 regex = "1.10"
 lazy_static = "1"
 semver = "1.0.25"
+crunchy = "0.2.4"
 # Temporary pins to work around downstream issues
 # https://github.com/apache/arrow-rs/commit/2fddf85afcd20110ce783ed5b4cdeb82293da30b
 chrono = "=0.4.41"
 # https://github.com/RustCrypto/formats/issues/1684
 base64ct = "=1.6.0"
-# Workaround for: https://github.com/eira-fransham/crunchy/issues/13
-crunchy = "=0.2.2"
 # Workaround for: https://github.com/Lokathor/bytemuck/issues/306
 bytemuck_derive = ">=1.8.1, <1.9.0"

ci/set_lance_version.py

@@ -1,4 +1,5 @@
 import argparse
+import re
 import sys
 import json
 
@@ -18,8 +19,12 @@ def run_command(command: str) -> str:
 
 def get_latest_stable_version() -> str:
     version_line = run_command("cargo info lance | grep '^version:'")
-    version = version_line.split(" ")[1].strip()
-    return version
+    # Example output: "version: 0.35.0 (latest 0.37.0)"
+    match = re.search(r'\(latest ([0-9.]+)\)', version_line)
+    if match:
+        return match.group(1)
+    # Fallback: use the first version after 'version:'
+    return version_line.split("version:")[1].split()[0].strip()
 
 
 def get_latest_preview_version() -> str:
@@ -54,6 +59,52 @@ def extract_features(line: str) -> list:
     return []
 
 
+def extract_default_features(line: str) -> bool:
+    """
+    Checks if default-features = false is present in a line in Cargo.toml.
+    Example: 'lance = { "version" = "=0.29.0", default-features = false, "features" = ["dynamodb"] }'
+    Returns: True if default-features = false is present, False otherwise
+    """
+    import re
+
+    match = re.search(r'default-features\s*=\s*false', line)
+    return match is not None
+
+
+def dict_to_toml_line(package_name: str, config: dict) -> str:
+    """
+    Converts a configuration dictionary to a TOML dependency line.
+    Dictionary insertion order is preserved (Python 3.7+), so the caller
+    controls the order of fields in the output.
+
+    Args:
+        package_name: The name of the package (e.g., "lance", "lance-io")
+        config: Dictionary with keys like "version", "path", "git", "tag", "features", "default-features"
+                The order of keys in this dict determines the order in the output.
+
+    Returns:
+        A properly formatted TOML line with a trailing newline
+    """
+    # If only version is specified, use simple format
+    if len(config) == 1 and "version" in config:
+        return f'{package_name} = "{config["version"]}"\n'
+
+    # Otherwise, use inline table format
+    parts = []
+    for key, value in config.items():
+        if key == "default-features" and not value:
+            parts.append("default-features = false")
+        elif key == "features":
+            parts.append(f'"features" = {json.dumps(value)}')
+        elif isinstance(value, str):
+            parts.append(f'"{key}" = "{value}"')
+        else:
+            # This shouldn't happen with our current usage
+            parts.append(f'"{key}" = {json.dumps(value)}')
+
+    return f'{package_name} = {{ {", ".join(parts)} }}\n'
+
+
 def update_cargo_toml(line_updater):
     """
     Updates the Cargo.toml file by applying the line_updater function to each line.
@@ -67,20 +118,27 @@ def update_cargo_toml(line_updater):
     is_parsing_lance_line = False
     for line in lines:
         if line.startswith("lance"):
-            # Update the line using the provided function
-            if line.strip().endswith("}"):
+            # Check if this is a single-line or multi-line entry
+            # Single-line entries either:
+            # 1. End with } (complete inline table)
+            # 2. End with " (simple version string)
+            # Multi-line entries start with { but don't end with }
+            if line.strip().endswith("}") or line.strip().endswith('"'):
+                # Single-line entry - process immediately
                 new_lines.append(line_updater(line))
-            else:
+            elif "{" in line and not line.strip().endswith("}"):
+                # Multi-line entry - start accumulating
                 lance_line = line
                 is_parsing_lance_line = True
+            else:
+                # Single-line entry without quotes or braces (shouldn't happen but handle it)
+                new_lines.append(line_updater(line))
         elif is_parsing_lance_line:
             lance_line += line
             if line.strip().endswith("}"):
                 new_lines.append(line_updater(lance_line))
                 lance_line = ""
                 is_parsing_lance_line = False
-            else:
-                print("doesn't end with }:", line)
         else:
             # Keep the line unchanged
             new_lines.append(line)
@@ -92,18 +150,25 @@ def update_cargo_toml(line_updater):
 def set_stable_version(version: str):
     """
     Sets lines to
-    lance = { "version" = "=0.29.0", "features" = ["dynamodb"] }
-    lance-io = "=0.29.0"
+    lance = { "version" = "=0.29.0", default-features = false, "features" = ["dynamodb"] }
+    lance-io = { "version" = "=0.29.0", default-features = false }
     ...
     """
 
     def line_updater(line: str) -> str:
         package_name = line.split("=", maxsplit=1)[0].strip()
+
+        # Build config in desired order: version, default-features, features
+        config = {"version": f"={version}"}
+
+        if extract_default_features(line):
+            config["default-features"] = False
+
         features = extract_features(line)
         if features:
-            return f'{package_name} = {{ "version" = "={version}", "features" = {json.dumps(features)} }}\n'
-        else:
-            return f'{package_name} = "={version}"\n'
+            config["features"] = features
+
+        return dict_to_toml_line(package_name, config)
 
     update_cargo_toml(line_updater)
 
@@ -111,19 +176,29 @@ def set_stable_version(version: str):
 def set_preview_version(version: str):
     """
     Sets lines to
-    lance = { "version" = "=0.29.0", "features" = ["dynamodb"], tag = "v0.29.0-beta.2", git="https://github.com/lancedb/lance.git" }
-    lance-io = { version = "=0.29.0", tag = "v0.29.0-beta.2", git="https://github.com/lancedb/lance.git" }
+    lance = { "version" = "=0.29.0", default-features = false, "features" = ["dynamodb"], "tag" = "v0.29.0-beta.2", "git" = "https://github.com/lancedb/lance.git" }
+    lance-io = { "version" = "=0.29.0", default-features = false, "tag" = "v0.29.0-beta.2", "git" = "https://github.com/lancedb/lance.git" }
     ...
     """
 
     def line_updater(line: str) -> str:
         package_name = line.split("=", maxsplit=1)[0].strip()
-        features = extract_features(line)
         base_version = version.split("-")[0]  # Get the base version without beta suffix
+
+        # Build config in desired order: version, default-features, features, tag, git
+        config = {"version": f"={base_version}"}
+
+        if extract_default_features(line):
+            config["default-features"] = False
+
+        features = extract_features(line)
         if features:
-            return f'{package_name} = {{ "version" = "={base_version}", "features" = {json.dumps(features)}, "tag" = "v{version}", "git" = "https://github.com/lancedb/lance.git" }}\n'
-        else:
-            return f'{package_name} = {{ "version" = "={base_version}", "tag" = "v{version}", "git" = "https://github.com/lancedb/lance.git" }}\n'
+            config["features"] = features
+
+        config["tag"] = f"v{version}"
+        config["git"] = "https://github.com/lancedb/lance.git"
+
+        return dict_to_toml_line(package_name, config)
 
     update_cargo_toml(line_updater)
 
@@ -131,18 +206,25 @@ def set_preview_version(version: str):
 def set_local_version():
     """
     Sets lines to
-    lance = { path = "../lance/rust/lance", features = ["dynamodb"] }
-    lance-io = { path = "../lance/rust/lance-io" }
+    lance = { "path" = "../lance/rust/lance", default-features = false, "features" = ["dynamodb"] }
+    lance-io = { "path" = "../lance/rust/lance-io", default-features = false }
     ...
     """
 
     def line_updater(line: str) -> str:
         package_name = line.split("=", maxsplit=1)[0].strip()
+
+        # Build config in desired order: path, default-features, features
+        config = {"path": f"../lance/rust/{package_name}"}
+
+        if extract_default_features(line):
+            config["default-features"] = False
+
         features = extract_features(line)
         if features:
-            return f'{package_name} = {{ "path" = "../lance/rust/{package_name}", "features" = {json.dumps(features)} }}\n'
-        else:
-            return f'{package_name} = {{ "path" = "../lance/rust/{package_name}" }}\n'
+            config["features"] = features
+
+        return dict_to_toml_line(package_name, config)
 
     update_cargo_toml(line_updater)
 

ci/update_lockfiles.sh

@@ -15,16 +15,13 @@ cargo metadata --quiet > /dev/null
 pushd nodejs || exit 1
 npm install --package-lock-only --silent
 popd
-pushd node || exit 1
-npm install --package-lock-only --silent
-popd
 
 if git diff --quiet --exit-code; then
   echo "No lockfile changes to commit; skipping amend."
 elif $AMEND; then
-  git add Cargo.lock nodejs/package-lock.json node/package-lock.json
+  git add Cargo.lock nodejs/package-lock.json
   git commit --amend --no-edit
 else
-  git add Cargo.lock nodejs/package-lock.json node/package-lock.json
+  git add Cargo.lock nodejs/package-lock.json
   git commit -m "Update lockfiles"
 fi

docs/src/concepts/data_management.md

@@ -13,7 +13,7 @@ The following concepts are important to keep in mind:
 - Data is versioned, with each insert operation creating a new version of the dataset and an update to the manifest that tracks versions via metadata
 
 !!! note
-    1. First, each version contains metadata and just the new/updated data in your transaction. So if you have 100 versions, they aren't 100 duplicates of the same data. However, they do have 100x the metadata overhead of a single version, which can result in slower queries.  
+    1. First, each version contains metadata and just the new/updated data in your transaction. So if you have 100 versions, they aren't 100 duplicates of the same data. However, they do have 100x the metadata overhead of a single version, which can result in slower queries.
     2. Second, these versions exist to keep LanceDB scalable and consistent. We do not immediately blow away old versions when creating new ones because other clients might be in the middle of querying the old version. It's important to retain older versions for as long as they might be queried.
 
 ## What are fragments?
@@ -37,6 +37,10 @@ Depending on the use case and dataset, optimal compaction will have different re
 - It’s always better to use *batch* inserts rather than adding 1 row at a time (to avoid too small fragments). If single-row inserts are unavoidable, run compaction on a regular basis to merge them into larger fragments.
 - Keep the number of fragments under 100, which is suitable for most use cases (for *really* large datasets of >500M rows, more fragments might be needed)
 
+!!! note
+
+    LanceDB Cloud/Enterprise supports [auto-compaction](https://docs.lancedb.com/enterprise/architecture/architecture#write-path) which automatically optimizes fragments in the background as data changes.
+
 ## Deletion
 
 Although Lance allows you to delete rows from a dataset, it does not actually delete the data immediately. It simply marks the row as deleted in the `DataFile` that represents a fragment. For a given version of the dataset, each fragment can have up to one deletion file (if no rows were ever deleted from that fragment, it will not have a deletion file). This is important to keep in mind because it means that the data is still there, and can be recovered if needed, as long as that version still exists based on your backup policy.
@@ -50,13 +54,9 @@ Reindexing is the process of updating the index to account for new data, keeping
 
 Both LanceDB OSS and Cloud support reindexing, but the process (at least for now) is different for each, depending on the type of index.
 
-When a reindex job is triggered in the background, the entire data is reindexed, but in the interim as new queries come in, LanceDB will combine results from the existing index with exhaustive kNN search on the new data. This is done to ensure that you're still searching on all your data, but it does come at a performance cost. The more data that you add without reindexing, the impact on latency (due to exhaustive search) can be noticeable.
+In LanceDB OSS, re-indexing happens synchronously when you call either `create_index` or `optimize` on a table. In LanceDB Cloud, re-indexing happens asynchronously as you add and update data in your table.
 
-### Vector reindex
+By default, queries will search new data even if it has yet to be indexed. This is done using brute-force methods, such as kNN for vector search, and combined with the fast index search results. This is done to ensure that you're always searching over all your data, but it does come at a performance cost. Without reindexing, adding more data to a table will make queries slower and more expensive. This behavior can be disabled by setting the [fast_search](https://lancedb.github.io/lancedb/python/python/#lancedb.query.AsyncQuery.fast_search) parameter which will instruct the query to ignore un-indexed data.
 
-* LanceDB Cloud supports incremental reindexing, where a background process will trigger a new index build for you automatically when new data is added to a dataset
+* LanceDB Cloud/Enterprise supports [automatic incremental reindexing](https://docs.lancedb.com/core#vector-index) for vector, scalar, and FTS indices, where a background process will trigger a new index build for you automatically when new data is added or modified in a dataset
 * LanceDB OSS requires you to manually trigger a reindex operation -- we are working on adding incremental reindexing to LanceDB OSS as well
-
-### FTS reindex
-
-FTS reindexing is supported in both LanceDB OSS and Cloud, but requires that it's manually rebuilt once you have a significant enough amount of new data added that needs to be reindexed. We [updated](https://github.com/lancedb/lancedb/pull/762) Tantivy's default heap size from 128MB to 1GB in LanceDB to make it much faster to reindex, by up to 10x from the default settings.

docs/src/js/classes/Connection.md

@@ -45,6 +45,8 @@ Any attempt to use the connection after it is closed will result in an error.
 
 ### createEmptyTable()
 
+#### createEmptyTable(name, schema, options)
+
 ```ts
 abstract createEmptyTable(
    name,
@@ -54,7 +56,7 @@ abstract createEmptyTable(
 
 Creates a new empty Table
 
-#### Parameters
+##### Parameters
 
 * **name**: `string`
     The name of the table.
@@ -63,8 +65,39 @@ Creates a new empty Table
     The schema of the table
 
 * **options?**: `Partial`<[`CreateTableOptions`](../interfaces/CreateTableOptions.md)>
+    Additional options (backwards compatibility)
 
-#### Returns
+##### Returns
+
+`Promise`<[`Table`](Table.md)>
+
+#### createEmptyTable(name, schema, namespace, options)
+
+```ts
+abstract createEmptyTable(
+   name,
+   schema,
+   namespace?,
+   options?): Promise<Table>
+```
+
+Creates a new empty Table
+
+##### Parameters
+
+* **name**: `string`
+    The name of the table.
+
+* **schema**: [`SchemaLike`](../type-aliases/SchemaLike.md)
+    The schema of the table
+
+* **namespace?**: `string`[]
+    The namespace to create the table in (defaults to root namespace)
+
+* **options?**: `Partial`&lt;[`CreateTableOptions`](../interfaces/CreateTableOptions.md)&gt;
+    Additional options
+
+##### Returns
 
 `Promise`&lt;[`Table`](Table.md)&gt;
 
@@ -72,10 +105,10 @@ Creates a new empty Table
 
 ### createTable()
 
-#### createTable(options)
+#### createTable(options, namespace)
 
 ```ts
-abstract createTable(options): Promise<Table>
+abstract createTable(options, namespace?): Promise<Table>
 ```
 
 Creates a new Table and initialize it with new data.
@@ -85,6 +118,9 @@ Creates a new Table and initialize it with new data.
 * **options**: `object` & `Partial`&lt;[`CreateTableOptions`](../interfaces/CreateTableOptions.md)&gt;
     The options object.
 
+* **namespace?**: `string`[]
+    The namespace to create the table in (defaults to root namespace)
+
 ##### Returns
 
 `Promise`&lt;[`Table`](Table.md)&gt;
@@ -110,6 +146,38 @@ Creates a new Table and initialize it with new data.
     to be inserted into the table
 
 * **options?**: `Partial`&lt;[`CreateTableOptions`](../interfaces/CreateTableOptions.md)&gt;
+    Additional options (backwards compatibility)
+
+##### Returns
+
+`Promise`&lt;[`Table`](Table.md)&gt;
+
+#### createTable(name, data, namespace, options)
+
+```ts
+abstract createTable(
+   name,
+   data,
+   namespace?,
+   options?): Promise<Table>
+```
+
+Creates a new Table and initialize it with new data.
+
+##### Parameters
+
+* **name**: `string`
+    The name of the table.
+
+* **data**: [`TableLike`](../type-aliases/TableLike.md) \| `Record`&lt;`string`, `unknown`&gt;[]
+    Non-empty Array of Records
+    to be inserted into the table
+
+* **namespace?**: `string`[]
+    The namespace to create the table in (defaults to root namespace)
+
+* **options?**: `Partial`&lt;[`CreateTableOptions`](../interfaces/CreateTableOptions.md)&gt;
+    Additional options
 
 ##### Returns
 
@@ -134,11 +202,16 @@ Return a brief description of the connection
 ### dropAllTables()
 
 ```ts
-abstract dropAllTables(): Promise<void>
+abstract dropAllTables(namespace?): Promise<void>
 ```
 
 Drop all tables in the database.
 
+#### Parameters
+
+* **namespace?**: `string`[]
+    The namespace to drop tables from (defaults to root namespace).
+
 #### Returns
 
 `Promise`&lt;`void`&gt;
@@ -148,7 +221,7 @@ Drop all tables in the database.
 ### dropTable()
 
 ```ts
-abstract dropTable(name): Promise<void>
+abstract dropTable(name, namespace?): Promise<void>
 ```
 
 Drop an existing table.
@@ -158,6 +231,9 @@ Drop an existing table.
 * **name**: `string`
     The name of the table to drop.
 
+* **namespace?**: `string`[]
+    The namespace of the table (defaults to root namespace).
+
 #### Returns
 
 `Promise`&lt;`void`&gt;
@@ -181,7 +257,10 @@ Return true if the connection has not been closed
 ### openTable()
 
 ```ts
-abstract openTable(name, options?): Promise<Table>
+abstract openTable(
+   name,
+   namespace?,
+   options?): Promise<Table>
 ```
 
 Open a table in the database.
@@ -191,7 +270,11 @@ Open a table in the database.
 * **name**: `string`
     The name of the table
 
+* **namespace?**: `string`[]
+    The namespace of the table (defaults to root namespace)
+
 * **options?**: `Partial`&lt;[`OpenTableOptions`](../interfaces/OpenTableOptions.md)&gt;
+    Additional options
 
 #### Returns
 
@@ -201,6 +284,8 @@ Open a table in the database.
 
 ### tableNames()
 
+#### tableNames(options)
+
 ```ts
 abstract tableNames(options?): Promise<string[]>
 ```
@@ -209,12 +294,35 @@ List all the table names in this database.
 
 Tables will be returned in lexicographical order.
 
-#### Parameters
+##### Parameters
+
+* **options?**: `Partial`&lt;[`TableNamesOptions`](../interfaces/TableNamesOptions.md)&gt;
+    options to control the
+    paging / start point (backwards compatibility)
+
+##### Returns
+
+`Promise`&lt;`string`[]&gt;
+
+#### tableNames(namespace, options)
+
+```ts
+abstract tableNames(namespace?, options?): Promise<string[]>
+```
+
+List all the table names in this database.
+
+Tables will be returned in lexicographical order.
+
+##### Parameters
+
+* **namespace?**: `string`[]
+    The namespace to list tables from (defaults to root namespace)
 
 * **options?**: `Partial`&lt;[`TableNamesOptions`](../interfaces/TableNamesOptions.md)&gt;
     options to control the
     paging / start point
 
-#### Returns
+##### Returns
 
 `Promise`&lt;`string`[]&gt;

docs/src/js/classes/HeaderProvider.md

@@ -0,0 +1,85 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / HeaderProvider
+
+# Class: `abstract` HeaderProvider
+
+Abstract base class for providing custom headers for each request.
+
+Users can implement this interface to provide dynamic headers for various purposes
+such as authentication (OAuth tokens, API keys), request tracking (correlation IDs),
+custom metadata, or any other header-based requirements. The provider is called
+before each request to ensure fresh header values are always used.
+
+## Examples
+
+Simple JWT token provider:
+```typescript
+class JWTProvider extends HeaderProvider {
+  constructor(private token: string) {
+    super();
+  }
+
+  getHeaders(): Record<string, string> {
+    return { authorization: `Bearer ${this.token}` };
+  }
+}
+```
+
+Provider with request tracking:
+```typescript
+class RequestTrackingProvider extends HeaderProvider {
+  constructor(private sessionId: string) {
+    super();
+  }
+
+  getHeaders(): Record<string, string> {
+    return {
+      "X-Session-Id": this.sessionId,
+      "X-Request-Id": `req-${Date.now()}`
+    };
+  }
+}
+```
+
+## Extended by
+
+- [`StaticHeaderProvider`](StaticHeaderProvider.md)
+- [`OAuthHeaderProvider`](OAuthHeaderProvider.md)
+
+## Constructors
+
+### new HeaderProvider()
+
+```ts
+new HeaderProvider(): HeaderProvider
+```
+
+#### Returns
+
+[`HeaderProvider`](HeaderProvider.md)
+
+## Methods
+
+### getHeaders()
+
+```ts
+abstract getHeaders(): Record<string, string>
+```
+
+Get the latest headers to be added to requests.
+
+This method is called before each request to the remote LanceDB server.
+Implementations should return headers that will be merged with existing headers.
+
+#### Returns
+
+`Record`&lt;`string`, `string`&gt;
+
+Dictionary of header names to values to add to the request.
+
+#### Throws
+
+If unable to fetch headers, the exception will be propagated and the request will fail.

docs/src/js/classes/NativeJsHeaderProvider.md

@@ -0,0 +1,29 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / NativeJsHeaderProvider
+
+# Class: NativeJsHeaderProvider
+
+JavaScript HeaderProvider implementation that wraps a JavaScript callback.
+This is the only native header provider - all header provider implementations
+should provide a JavaScript function that returns headers.
+
+## Constructors
+
+### new NativeJsHeaderProvider()
+
+```ts
+new NativeJsHeaderProvider(getHeadersCallback): NativeJsHeaderProvider
+```
+
+Create a new JsHeaderProvider from a JavaScript callback
+
+#### Parameters
+
+* **getHeadersCallback**
+
+#### Returns
+
+[`NativeJsHeaderProvider`](NativeJsHeaderProvider.md)

docs/src/js/classes/OAuthHeaderProvider.md

@@ -0,0 +1,108 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / OAuthHeaderProvider
+
+# Class: OAuthHeaderProvider
+
+Example implementation: OAuth token provider with automatic refresh.
+
+This is an example implementation showing how to manage OAuth tokens
+with automatic refresh when they expire.
+
+## Example
+
+```typescript
+async function fetchToken(): Promise<TokenResponse> {
+  const response = await fetch("https://oauth.example.com/token", {
+    method: "POST",
+    body: JSON.stringify({
+      grant_type: "client_credentials",
+      client_id: "your-client-id",
+      client_secret: "your-client-secret"
+    }),
+    headers: { "Content-Type": "application/json" }
+  });
+  const data = await response.json();
+  return {
+    accessToken: data.access_token,
+    expiresIn: data.expires_in
+  };
+}
+
+const provider = new OAuthHeaderProvider(fetchToken);
+const headers = provider.getHeaders();
+// Returns: {"authorization": "Bearer <your-token>"}
+```
+
+## Extends
+
+- [`HeaderProvider`](HeaderProvider.md)
+
+## Constructors
+
+### new OAuthHeaderProvider()
+
+```ts
+new OAuthHeaderProvider(tokenFetcher, refreshBufferSeconds): OAuthHeaderProvider
+```
+
+Initialize the OAuth provider.
+
+#### Parameters
+
+* **tokenFetcher**
+    Function to fetch new tokens. Should return object with 'accessToken' and optionally 'expiresIn'.
+
+* **refreshBufferSeconds**: `number` = `300`
+    Seconds before expiry to refresh token. Default 300 (5 minutes).
+
+#### Returns
+
+[`OAuthHeaderProvider`](OAuthHeaderProvider.md)
+
+#### Overrides
+
+[`HeaderProvider`](HeaderProvider.md).[`constructor`](HeaderProvider.md#constructors)
+
+## Methods
+
+### getHeaders()
+
+```ts
+getHeaders(): Record<string, string>
+```
+
+Get OAuth headers, refreshing token if needed.
+Note: This is synchronous for now as the Rust implementation expects sync.
+In a real implementation, this would need to handle async properly.
+
+#### Returns
+
+`Record`&lt;`string`, `string`&gt;
+
+Headers with Bearer token authorization.
+
+#### Throws
+
+If unable to fetch or refresh token.
+
+#### Overrides
+
+[`HeaderProvider`](HeaderProvider.md).[`getHeaders`](HeaderProvider.md#getheaders)
+
+***
+
+### refreshToken()
+
+```ts
+refreshToken(): Promise<void>
+```
+
+Manually refresh the token.
+Call this before using getHeaders() to ensure token is available.
+
+#### Returns
+
+`Promise`&lt;`void`&gt;

docs/src/js/classes/Query.md

@@ -14,7 +14,7 @@ A builder for LanceDB queries.
 
 ## Extends
 
-- [`QueryBase`](QueryBase.md)<`NativeQuery`>
+- `StandardQueryBase`<`NativeQuery`>
 
 ## Properties
 
@@ -26,7 +26,7 @@ protected inner: Query | Promise<Query>;
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`inner`](QueryBase.md#inner)
+`StandardQueryBase.inner`
 
 ## Methods
 
@@ -73,7 +73,7 @@ AnalyzeExec verbose=true, metrics=[]
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`analyzePlan`](QueryBase.md#analyzeplan)
+`StandardQueryBase.analyzePlan`
 
 ***
 
@@ -107,7 +107,7 @@ single query)
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`execute`](QueryBase.md#execute)
+`StandardQueryBase.execute`
 
 ***
 
@@ -143,7 +143,7 @@ const plan = await table.query().nearestTo([0.5, 0.2]).explainPlan();
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`explainPlan`](QueryBase.md#explainplan)
+`StandardQueryBase.explainPlan`
 
 ***
 
@@ -164,7 +164,7 @@ Use [Table#optimize](Table.md#optimize) to index all un-indexed data.
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`fastSearch`](QueryBase.md#fastsearch)
+`StandardQueryBase.fastSearch`
 
 ***
 
@@ -194,7 +194,7 @@ Use `where` instead
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`filter`](QueryBase.md#filter)
+`StandardQueryBase.filter`
 
 ***
 
@@ -216,7 +216,7 @@ fullTextSearch(query, options?): this
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`fullTextSearch`](QueryBase.md#fulltextsearch)
+`StandardQueryBase.fullTextSearch`
 
 ***
 
@@ -241,7 +241,7 @@ called then every valid row from the table will be returned.
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`limit`](QueryBase.md#limit)
+`StandardQueryBase.limit`
 
 ***
 
@@ -325,6 +325,10 @@ nearestToText(query, columns?): Query
 offset(offset): this
 ```
 
+Set the number of rows to skip before returning results.
+
+This is useful for pagination.
+
 #### Parameters
 
 * **offset**: `number`
@@ -335,7 +339,7 @@ offset(offset): this
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`offset`](QueryBase.md#offset)
+`StandardQueryBase.offset`
 
 ***
 
@@ -388,7 +392,7 @@ object insertion order is easy to get wrong and `Map` is more foolproof.
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`select`](QueryBase.md#select)
+`StandardQueryBase.select`
 
 ***
 
@@ -410,7 +414,7 @@ Collect the results as an array of objects.
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`toArray`](QueryBase.md#toarray)
+`StandardQueryBase.toArray`
 
 ***
 
@@ -436,7 +440,7 @@ ArrowTable.
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`toArrow`](QueryBase.md#toarrow)
+`StandardQueryBase.toArrow`
 
 ***
 
@@ -471,7 +475,7 @@ on the filter column(s).
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`where`](QueryBase.md#where)
+`StandardQueryBase.where`
 
 ***
 
@@ -493,4 +497,4 @@ order to perform hybrid search.
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`withRowId`](QueryBase.md#withrowid)
+`StandardQueryBase.withRowId`

docs/src/js/classes/QueryBase.md

@@ -15,12 +15,11 @@ Common methods supported by all query types
 
 ## Extended by
 
-- [`Query`](Query.md)
-- [`VectorQuery`](VectorQuery.md)
+- [`TakeQuery`](TakeQuery.md)
 
 ## Type Parameters
 
-• **NativeQueryType** *extends* `NativeQuery` \| `NativeVectorQuery`
+• **NativeQueryType** *extends* `NativeQuery` \| `NativeVectorQuery` \| `NativeTakeQuery`
 
 ## Implements
 
@@ -141,104 +140,6 @@ const plan = await table.query().nearestTo([0.5, 0.2]).explainPlan();
 
 ***
 
-### fastSearch()
-
-```ts
-fastSearch(): this
-```
-
-Skip searching un-indexed data. This can make search faster, but will miss
-any data that is not yet indexed.
-
-Use [Table#optimize](Table.md#optimize) to index all un-indexed data.
-
-#### Returns
-
-`this`
-
-***
-
-### ~~filter()~~
-
-```ts
-filter(predicate): this
-```
-
-A filter statement to be applied to this query.
-
-#### Parameters
-
-* **predicate**: `string`
-
-#### Returns
-
-`this`
-
-#### See
-
-where
-
-#### Deprecated
-
-Use `where` instead
-
-***
-
-### fullTextSearch()
-
-```ts
-fullTextSearch(query, options?): this
-```
-
-#### Parameters
-
-* **query**: `string` \| [`FullTextQuery`](../interfaces/FullTextQuery.md)
-
-* **options?**: `Partial`<[`FullTextSearchOptions`](../interfaces/FullTextSearchOptions.md)>
-
-#### Returns
-
-`this`
-
-***
-
-### limit()
-
-```ts
-limit(limit): this
-```
-
-Set the maximum number of results to return.
-
-By default, a plain search has no limit.  If this method is not
-called then every valid row from the table will be returned.
-
-#### Parameters
-
-* **limit**: `number`
-
-#### Returns
-
-`this`
-
-***
-
-### offset()
-
-```ts
-offset(offset): this
-```
-
-#### Parameters
-
-* **offset**: `number`
-
-#### Returns
-
-`this`
-
-***
-
 ### select()
 
 ```ts
@@ -328,37 +229,6 @@ ArrowTable.
 
 ***
 
-### where()
-
-```ts
-where(predicate): this
-```
-
-A filter statement to be applied to this query.
-
-The filter should be supplied as an SQL query string.  For example:
-
-#### Parameters
-
-* **predicate**: `string`
-
-#### Returns
-
-`this`
-
-#### Example
-
-```ts
-x > 10
-y > 0 AND y < 100
-x > 5 OR y = 'test'
-
-Filtering performance can often be improved by creating a scalar index
-on the filter column(s).
-```
-
-***
-
 ### withRowId()
 
 ```ts

docs/src/js/classes/Session.md

@@ -9,7 +9,8 @@
 A session for managing caches and object stores across LanceDB operations.
 
 Sessions allow you to configure cache sizes for index and metadata caches,
-which can significantly impact performance for large datasets.
+which can significantly impact memory use and performance. They can
+also be re-used across multiple connections to share the same cache state.
 
 ## Constructors
 
@@ -24,8 +25,11 @@ Create a new session with custom cache sizes.
 # Parameters
 
 - `index_cache_size_bytes`: The size of the index cache in bytes.
+  Index data is stored in memory in this cache to speed up queries.
   Defaults to 6GB if not specified.
 - `metadata_cache_size_bytes`: The size of the metadata cache in bytes.
+  The metadata cache stores file metadata and schema information in memory.
+  This cache improves scan and write performance.
   Defaults to 1GB if not specified.
 
 #### Parameters

docs/src/js/classes/StaticHeaderProvider.md

@@ -0,0 +1,70 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / StaticHeaderProvider
+
+# Class: StaticHeaderProvider
+
+Example implementation: A simple header provider that returns static headers.
+
+This is an example implementation showing how to create a HeaderProvider
+for cases where headers don't change during the session.
+
+## Example
+
+```typescript
+const provider = new StaticHeaderProvider({
+  authorization: "Bearer my-token",
+  "X-Custom-Header": "custom-value"
+});
+const headers = provider.getHeaders();
+// Returns: {authorization: 'Bearer my-token', 'X-Custom-Header': 'custom-value'}
+```
+
+## Extends
+
+- [`HeaderProvider`](HeaderProvider.md)
+
+## Constructors
+
+### new StaticHeaderProvider()
+
+```ts
+new StaticHeaderProvider(headers): StaticHeaderProvider
+```
+
+Initialize with static headers.
+
+#### Parameters
+
+* **headers**: `Record`<`string`, `string`>
+    Headers to return for every request.
+
+#### Returns
+
+[`StaticHeaderProvider`](StaticHeaderProvider.md)
+
+#### Overrides
+
+[`HeaderProvider`](HeaderProvider.md).[`constructor`](HeaderProvider.md#constructors)
+
+## Methods
+
+### getHeaders()
+
+```ts
+getHeaders(): Record<string, string>
+```
+
+Return the static headers.
+
+#### Returns
+
+`Record`&lt;`string`, `string`&gt;
+
+Copy of the static headers.
+
+#### Overrides
+
+[`HeaderProvider`](HeaderProvider.md).[`getHeaders`](HeaderProvider.md#getheaders)

docs/src/js/classes/Table.md

@@ -674,6 +674,48 @@ console.log(tags); // { "v1": { version: 1, manifestSize: ... } }
 
 ***
 
+### takeOffsets()
+
+```ts
+abstract takeOffsets(offsets): TakeQuery
+```
+
+Create a query that returns a subset of the rows in the table.
+
+#### Parameters
+
+* **offsets**: `number`[]
+    The offsets of the rows to return.
+
+#### Returns
+
+[`TakeQuery`](TakeQuery.md)
+
+A builder that can be used to parameterize the query.
+
+***
+
+### takeRowIds()
+
+```ts
+abstract takeRowIds(rowIds): TakeQuery
+```
+
+Create a query that returns a subset of the rows in the table.
+
+#### Parameters
+
+* **rowIds**: `number`[]
+    The row ids of the rows to return.
+
+#### Returns
+
+[`TakeQuery`](TakeQuery.md)
+
+A builder that can be used to parameterize the query.
+
+***
+
 ### toArrow()
 
 ```ts

docs/src/js/classes/TakeQuery.md

@@ -0,0 +1,265 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / TakeQuery
+
+# Class: TakeQuery
+
+A query that returns a subset of the rows in the table.
+
+## Extends
+
+- [`QueryBase`](QueryBase.md)<`NativeTakeQuery`>
+
+## Properties
+
+### inner
+
+```ts
+protected inner: TakeQuery | Promise<TakeQuery>;
+```
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`inner`](QueryBase.md#inner)
+
+## Methods
+
+### analyzePlan()
+
+```ts
+analyzePlan(): Promise<string>
+```
+
+Executes the query and returns the physical query plan annotated with runtime metrics.
+
+This is useful for debugging and performance analysis, as it shows how the query was executed
+and includes metrics such as elapsed time, rows processed, and I/O statistics.
+
+#### Returns
+
+`Promise`&lt;`string`&gt;
+
+A query execution plan with runtime metrics for each step.
+
+#### Example
+
+```ts
+import * as lancedb from "@lancedb/lancedb"
+
+const db = await lancedb.connect("./.lancedb");
+const table = await db.createTable("my_table", [
+  { vector: [1.1, 0.9], id: "1" },
+]);
+
+const plan = await table.query().nearestTo([0.5, 0.2]).analyzePlan();
+
+Example output (with runtime metrics inlined):
+AnalyzeExec verbose=true, metrics=[]
+ ProjectionExec: expr=[id@3 as id, vector@0 as vector, _distance@2 as _distance], metrics=[output_rows=1, elapsed_compute=3.292µs]
+  Take: columns="vector, _rowid, _distance, (id)", metrics=[output_rows=1, elapsed_compute=66.001µs, batches_processed=1, bytes_read=8, iops=1, requests=1]
+   CoalesceBatchesExec: target_batch_size=1024, metrics=[output_rows=1, elapsed_compute=3.333µs]
+    GlobalLimitExec: skip=0, fetch=10, metrics=[output_rows=1, elapsed_compute=167ns]
+     FilterExec: _distance@2 IS NOT NULL, metrics=[output_rows=1, elapsed_compute=8.542µs]
+      SortExec: TopK(fetch=10), expr=[_distance@2 ASC NULLS LAST], metrics=[output_rows=1, elapsed_compute=63.25µs, row_replacements=1]
+       KNNVectorDistance: metric=l2, metrics=[output_rows=1, elapsed_compute=114.333µs, output_batches=1]
+        LanceScan: uri=/path/to/data, projection=[vector], row_id=true, row_addr=false, ordered=false, metrics=[output_rows=1, elapsed_compute=103.626µs, bytes_read=549, iops=2, requests=2]
+```
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`analyzePlan`](QueryBase.md#analyzeplan)
+
+***
+
+### execute()
+
+```ts
+protected execute(options?): RecordBatchIterator
+```
+
+Execute the query and return the results as an
+
+#### Parameters
+
+* **options?**: `Partial`&lt;[`QueryExecutionOptions`](../interfaces/QueryExecutionOptions.md)&gt;
+
+#### Returns
+
+[`RecordBatchIterator`](RecordBatchIterator.md)
+
+#### See
+
+ - AsyncIterator
+of
+ - RecordBatch.
+
+By default, LanceDb will use many threads to calculate results and, when
+the result set is large, multiple batches will be processed at one time.
+This readahead is limited however and backpressure will be applied if this
+stream is consumed slowly (this constrains the maximum memory used by a
+single query)
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`execute`](QueryBase.md#execute)
+
+***
+
+### explainPlan()
+
+```ts
+explainPlan(verbose): Promise<string>
+```
+
+Generates an explanation of the query execution plan.
+
+#### Parameters
+
+* **verbose**: `boolean` = `false`
+    If true, provides a more detailed explanation. Defaults to false.
+
+#### Returns
+
+`Promise`&lt;`string`&gt;
+
+A Promise that resolves to a string containing the query execution plan explanation.
+
+#### Example
+
+```ts
+import * as lancedb from "@lancedb/lancedb"
+const db = await lancedb.connect("./.lancedb");
+const table = await db.createTable("my_table", [
+  { vector: [1.1, 0.9], id: "1" },
+]);
+const plan = await table.query().nearestTo([0.5, 0.2]).explainPlan();
+```
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`explainPlan`](QueryBase.md#explainplan)
+
+***
+
+### select()
+
+```ts
+select(columns): this
+```
+
+Return only the specified columns.
+
+By default a query will return all columns from the table.  However, this can have
+a very significant impact on latency.  LanceDb stores data in a columnar fashion.  This
+means we can finely tune our I/O to select exactly the columns we need.
+
+As a best practice you should always limit queries to the columns that you need.  If you
+pass in an array of column names then only those columns will be returned.
+
+You can also use this method to create new "dynamic" columns based on your existing columns.
+For example, you may not care about "a" or "b" but instead simply want "a + b".  This is often
+seen in the SELECT clause of an SQL query (e.g. `SELECT a+b FROM my_table`).
+
+To create dynamic columns you can pass in a Map<string, string>.  A column will be returned
+for each entry in the map.  The key provides the name of the column.  The value is
+an SQL string used to specify how the column is calculated.
+
+For example, an SQL query might state `SELECT a + b AS combined, c`.  The equivalent
+input to this method would be:
+
+#### Parameters
+
+* **columns**: `string` \| `string`[] \| `Record`&lt;`string`, `string`&gt; \| `Map`&lt;`string`, `string`&gt;
+
+#### Returns
+
+`this`
+
+#### Example
+
+```ts
+new Map([["combined", "a + b"], ["c", "c"]])
+
+Columns will always be returned in the order given, even if that order is different than
+the order used when adding the data.
+
+Note that you can pass in a `Record<string, string>` (e.g. an object literal). This method
+uses `Object.entries` which should preserve the insertion order of the object.  However,
+object insertion order is easy to get wrong and `Map` is more foolproof.
+```
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`select`](QueryBase.md#select)
+
+***
+
+### toArray()
+
+```ts
+toArray(options?): Promise<any[]>
+```
+
+Collect the results as an array of objects.
+
+#### Parameters
+
+* **options?**: `Partial`&lt;[`QueryExecutionOptions`](../interfaces/QueryExecutionOptions.md)&gt;
+
+#### Returns
+
+`Promise`&lt;`any`[]&gt;
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`toArray`](QueryBase.md#toarray)
+
+***
+
+### toArrow()
+
+```ts
+toArrow(options?): Promise<Table<any>>
+```
+
+Collect the results as an Arrow
+
+#### Parameters
+
+* **options?**: `Partial`&lt;[`QueryExecutionOptions`](../interfaces/QueryExecutionOptions.md)&gt;
+
+#### Returns
+
+`Promise`&lt;`Table`&lt;`any`&gt;&gt;
+
+#### See
+
+ArrowTable.
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`toArrow`](QueryBase.md#toarrow)
+
+***
+
+### withRowId()
+
+```ts
+withRowId(): this
+```
+
+Whether to return the row id in the results.
+
+This column can be used to match results between different queries. For
+example, to match results from a full text search and a vector search in
+order to perform hybrid search.
+
+#### Returns
+
+`this`
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`withRowId`](QueryBase.md#withrowid)

docs/src/js/classes/VectorQuery.md

@@ -16,7 +16,7 @@ This builder can be reused to execute the query many times.
 
 ## Extends
 
-- [`QueryBase`](QueryBase.md)<`NativeVectorQuery`>
+- `StandardQueryBase`<`NativeVectorQuery`>
 
 ## Properties
 
@@ -28,7 +28,7 @@ protected inner: VectorQuery | Promise<VectorQuery>;
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`inner`](QueryBase.md#inner)
+`StandardQueryBase.inner`
 
 ## Methods
 
@@ -91,7 +91,7 @@ AnalyzeExec verbose=true, metrics=[]
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`analyzePlan`](QueryBase.md#analyzeplan)
+`StandardQueryBase.analyzePlan`
 
 ***
 
@@ -248,7 +248,7 @@ single query)
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`execute`](QueryBase.md#execute)
+`StandardQueryBase.execute`
 
 ***
 
@@ -284,7 +284,7 @@ const plan = await table.query().nearestTo([0.5, 0.2]).explainPlan();
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`explainPlan`](QueryBase.md#explainplan)
+`StandardQueryBase.explainPlan`
 
 ***
 
@@ -305,7 +305,7 @@ Use [Table#optimize](Table.md#optimize) to index all un-indexed data.
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`fastSearch`](QueryBase.md#fastsearch)
+`StandardQueryBase.fastSearch`
 
 ***
 
@@ -335,7 +335,7 @@ Use `where` instead
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`filter`](QueryBase.md#filter)
+`StandardQueryBase.filter`
 
 ***
 
@@ -357,7 +357,7 @@ fullTextSearch(query, options?): this
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`fullTextSearch`](QueryBase.md#fulltextsearch)
+`StandardQueryBase.fullTextSearch`
 
 ***
 
@@ -382,7 +382,7 @@ called then every valid row from the table will be returned.
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`limit`](QueryBase.md#limit)
+`StandardQueryBase.limit`
 
 ***
 
@@ -480,6 +480,10 @@ the minimum and maximum to the same value.
 offset(offset): this
 ```
 
+Set the number of rows to skip before returning results.
+
+This is useful for pagination.
+
 #### Parameters
 
 * **offset**: `number`
@@ -490,7 +494,7 @@ offset(offset): this
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`offset`](QueryBase.md#offset)
+`StandardQueryBase.offset`
 
 ***
 
@@ -637,7 +641,7 @@ object insertion order is easy to get wrong and `Map` is more foolproof.
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`select`](QueryBase.md#select)
+`StandardQueryBase.select`
 
 ***
 
@@ -659,7 +663,7 @@ Collect the results as an array of objects.
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`toArray`](QueryBase.md#toarray)
+`StandardQueryBase.toArray`
 
 ***
 
@@ -685,7 +689,7 @@ ArrowTable.
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`toArrow`](QueryBase.md#toarrow)
+`StandardQueryBase.toArrow`
 
 ***
 
@@ -720,7 +724,7 @@ on the filter column(s).
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`where`](QueryBase.md#where)
+`StandardQueryBase.where`
 
 ***
 
@@ -742,4 +746,4 @@ order to perform hybrid search.
 
 #### Inherited from
 
-[`QueryBase`](QueryBase.md).[`withRowId`](QueryBase.md#withrowid)
+`StandardQueryBase.withRowId`

docs/src/js/functions/connect.md

@@ -6,13 +6,14 @@
 
 # Function: connect()
 
-## connect(uri, options, session)
+## connect(uri, options, session, headerProvider)
 
 ```ts
 function connect(
    uri,
    options?,
-   session?): Promise<Connection>
+   session?,
+   headerProvider?): Promise<Connection>
 ```
 
 Connect to a LanceDB instance at the given URI.
@@ -34,6 +35,8 @@ Accepted formats:
 
 * **session?**: [`Session`](../classes/Session.md)
 
+* **headerProvider?**: [`HeaderProvider`](../classes/HeaderProvider.md) \| () => `Record`&lt;`string`, `string`&gt; \| () => `Promise`&lt;`Record`&lt;`string`, `string`&gt;&gt;
+
 ### Returns
 
 `Promise`&lt;[`Connection`](../classes/Connection.md)&gt;
@@ -55,6 +58,18 @@ const conn = await connect(
 });
 ```
 
+Using with a header provider for per-request authentication:
+```ts
+const provider = new StaticHeaderProvider({
+  "X-API-Key": "my-key"
+});
+const conn = await connectWithHeaderProvider(
+  "db://host:port",
+  options,
+  provider
+);
+```
+
 ## connect(options)
 
 ```ts

docs/src/js/globals.md

@@ -20,19 +20,24 @@
 - [BooleanQuery](classes/BooleanQuery.md)
 - [BoostQuery](classes/BoostQuery.md)
 - [Connection](classes/Connection.md)
+- [HeaderProvider](classes/HeaderProvider.md)
 - [Index](classes/Index.md)
 - [MakeArrowTableOptions](classes/MakeArrowTableOptions.md)
 - [MatchQuery](classes/MatchQuery.md)
 - [MergeInsertBuilder](classes/MergeInsertBuilder.md)
 - [MultiMatchQuery](classes/MultiMatchQuery.md)
+- [NativeJsHeaderProvider](classes/NativeJsHeaderProvider.md)
+- [OAuthHeaderProvider](classes/OAuthHeaderProvider.md)
 - [PhraseQuery](classes/PhraseQuery.md)
 - [Query](classes/Query.md)
 - [QueryBase](classes/QueryBase.md)
 - [RecordBatchIterator](classes/RecordBatchIterator.md)
 - [Session](classes/Session.md)
+- [StaticHeaderProvider](classes/StaticHeaderProvider.md)
 - [Table](classes/Table.md)
 - [TagContents](classes/TagContents.md)
 - [Tags](classes/Tags.md)
+- [TakeQuery](classes/TakeQuery.md)
 - [VectorColumnOptions](classes/VectorColumnOptions.md)
 - [VectorQuery](classes/VectorQuery.md)
 
@@ -73,6 +78,7 @@
 - [TableNamesOptions](interfaces/TableNamesOptions.md)
 - [TableStatistics](interfaces/TableStatistics.md)
 - [TimeoutConfig](interfaces/TimeoutConfig.md)
+- [TokenResponse](interfaces/TokenResponse.md)
 - [UpdateOptions](interfaces/UpdateOptions.md)
 - [UpdateResult](interfaces/UpdateResult.md)
 - [Version](interfaces/Version.md)

docs/src/js/interfaces/ClientConfig.md

@@ -16,6 +16,14 @@ optional extraHeaders: Record<string, string>;
 
 ***
 
+### idDelimiter?
+
+```ts
+optional idDelimiter: string;
+```
+
+***
+
 ### retryConfig?
 
 ```ts

docs/src/js/interfaces/IndexOptions.md

@@ -26,6 +26,18 @@ will be used to determine the most useful kind of index to create.
 
 ***
 
+### name?
+
+```ts
+optional name: string;
+```
+
+Optional custom name for the index.
+
+If not provided, a default name will be generated based on the column name.
+
+***
+
 ### replace?
 
 ```ts
@@ -42,8 +54,27 @@ The default is true
 
 ***
 
+### train?
+
+```ts
+optional train: boolean;
+```
+
+Whether to train the index with existing data.
+
+If true (default), the index will be trained with existing data in the table.
+If false, the index will be created empty and populated as new data is added.
+
+Note: This option is only supported for scalar indices. Vector indices always train.
+
+***
+
 ### waitTimeoutSeconds?
 
 ```ts
 optional waitTimeoutSeconds: number;
 ```
+
+Timeout in seconds to wait for index creation to complete.
+
+If not specified, the method will return immediately after starting the index creation.

docs/src/js/interfaces/TimeoutConfig.md

@@ -44,3 +44,17 @@ optional readTimeout: number;
 The timeout for reading data from the server in seconds. Default is 300
 seconds (5 minutes). This can also be set via the environment variable
 `LANCE_CLIENT_READ_TIMEOUT`, as an integer number of seconds.
+
+***
+
+### timeout?
+
+```ts
+optional timeout: number;
+```
+
+The overall timeout for the entire request in seconds. This includes
+connection, send, and read time. If the entire request doesn't complete
+within this time, it will fail. Default is None (no overall timeout).
+This can also be set via the environment variable `LANCE_CLIENT_TIMEOUT`,
+as an integer number of seconds.

docs/src/js/interfaces/TokenResponse.md

@@ -0,0 +1,25 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / TokenResponse
+
+# Interface: TokenResponse
+
+Token response from OAuth provider.
+
+## Properties
+
+### accessToken
+
+```ts
+accessToken: string;
+```
+
+***
+
+### expiresIn?
+
+```ts
+optional expiresIn: number;
+```

java/core/lancedb-jni/Cargo.toml

@@ -15,7 +15,7 @@ publish = false
 crate-type = ["cdylib"]
 
 [dependencies]
-lancedb = { path = "../../../rust/lancedb" }
+lancedb = { path = "../../../rust/lancedb", default-features = false }
 lance = { workspace = true }
 arrow = { workspace = true, features = ["ffi"] }
 arrow-schema.workspace = true
@@ -25,3 +25,6 @@ snafu.workspace = true
 lazy_static.workspace = true
 serde = { version = "^1" }
 serde_json = { version = "1" }
+
+[features]
+default = ["lancedb/default"]

java/core/lancedb-jni/src/ffi.rs

@@ -16,6 +16,7 @@ pub trait JNIEnvExt {
     fn get_integers(&mut self, obj: &JObject) -> Result<Vec<i32>>;
 
     /// Get strings from Java List<String> object.
+    #[allow(dead_code)]
     fn get_strings(&mut self, obj: &JObject) -> Result<Vec<String>>;
 
     /// Get strings from Java String[] object.

java/core/lancedb-jni/src/traits.rs

@@ -6,6 +6,7 @@ use jni::JNIEnv;
 
 use crate::Result;
 
+#[allow(dead_code)]
 pub trait FromJObject<T> {
     fn extract(&self) -> Result<T>;
 }
@@ -39,6 +40,7 @@ impl FromJObject<f64> for JObject<'_> {
     }
 }
 
+#[allow(dead_code)]
 pub trait FromJString {
     fn extract(&self, env: &mut JNIEnv) -> Result<String>;
 }
@@ -66,6 +68,7 @@ pub trait JMapExt {
     fn get_f64(&self, env: &mut JNIEnv, key: &str) -> Result<Option<f64>>;
 }
 
+#[allow(dead_code)]
 fn get_map_value<T>(env: &mut JNIEnv, map: &JMap, key: &str) -> Result<Option<T>>
 where
     for<'a> JObject<'a>: FromJObject<T>,

java/core/pom.xml

@@ -8,7 +8,7 @@
     <parent>
         <groupId>com.lancedb</groupId>
         <artifactId>lancedb-parent</artifactId>
-        <version>0.21.2-final.0</version>
+        <version>0.22.1-beta.3</version>
         <relativePath>../pom.xml</relativePath>
     </parent>
 

java/lance-namespace/pom.xml

@@ -8,7 +8,7 @@
     <parent>
         <groupId>com.lancedb</groupId>
         <artifactId>lancedb-parent</artifactId>
-        <version>0.21.2-final.0</version>
+        <version>0.22.1-beta.3</version>
         <relativePath>../pom.xml</relativePath>
     </parent>
 

java/pom.xml

@@ -6,7 +6,7 @@
 
     <groupId>com.lancedb</groupId>
     <artifactId>lancedb-parent</artifactId>
-    <version>0.21.2-final.0</version>
+    <version>0.22.1-beta.3</version>
     <packaging>pom</packaging>
     <name>${project.artifactId}</name>
     <description>LanceDB Java SDK Parent POM</description>

nodejs/Cargo.toml

@@ -1,7 +1,7 @@
 [package]
 name = "lancedb-nodejs"
 edition.workspace = true
-version = "0.21.2"
+version = "0.22.1-beta.3"
 license.workspace = true
 description.workspace = true
 repository.workspace = true
@@ -18,7 +18,7 @@ arrow-array.workspace = true
 arrow-schema.workspace = true
 env_logger.workspace = true
 futures.workspace = true
-lancedb = { path = "../rust/lancedb" }
+lancedb = { path = "../rust/lancedb", default-features = false }
 napi = { version = "2.16.8", default-features = false, features = [
     "napi9",
     "async"
@@ -36,6 +36,6 @@ aws-lc-rs = "=1.13.0"
 napi-build = "2.1"
 
 [features]
-default = ["remote"]
+default = ["remote", "lancedb/default"]
 fp16kernels = ["lancedb/fp16kernels"]
 remote = ["lancedb/remote"]

nodejs/__test__/arrow.test.ts

@@ -1008,5 +1008,64 @@ describe.each([arrow15, arrow16, arrow17, arrow18])(
         expect(result).toEqual(null);
       });
     });
+
+    describe("boolean null handling", function () {
+      it("should handle null values in nullable boolean fields", () => {
+        const { makeArrowTable } = require("../lancedb/arrow");
+        const schema = new Schema([new Field("test", new arrow.Bool(), true)]);
+
+        // Test with all null values
+        const data = [{ test: null }];
+        const table = makeArrowTable(data, { schema });
+
+        expect(table.numRows).toBe(1);
+        expect(table.schema.names).toEqual(["test"]);
+        expect(table.getChild("test")!.get(0)).toBeNull();
+      });
+
+      it("should handle mixed null and non-null boolean values", () => {
+        const { makeArrowTable } = require("../lancedb/arrow");
+        const schema = new Schema([new Field("test", new Bool(), true)]);
+
+        // Test with mixed values
+        const data = [{ test: true }, { test: null }, { test: false }];
+        const table = makeArrowTable(data, { schema });
+
+        expect(table.numRows).toBe(3);
+        expect(table.getChild("test")!.get(0)).toBe(true);
+        expect(table.getChild("test")!.get(1)).toBeNull();
+        expect(table.getChild("test")!.get(2)).toBe(false);
+      });
+    });
   },
 );
+
+// Test for the undefined values bug fix
+describe("undefined values handling", () => {
+  it("should handle mixed undefined and actual values", () => {
+    const schema = new Schema([
+      new Field("text", new Utf8(), true), // nullable
+      new Field("number", new Int32(), true), // nullable
+      new Field("bool", new Bool(), true), // nullable
+    ]);
+
+    const data = [
+      { text: undefined, number: 42, bool: true },
+      { text: "hello", number: undefined, bool: false },
+      { text: "world", number: 123, bool: undefined },
+    ];
+    const table = makeArrowTable(data, { schema });
+
+    const result = table.toArray();
+    expect(result).toHaveLength(3);
+    expect(result[0].text).toBe(null);
+    expect(result[0].number).toBe(42);
+    expect(result[0].bool).toBe(true);
+    expect(result[1].text).toBe("hello");
+    expect(result[1].number).toBe(null);
+    expect(result[1].bool).toBe(false);
+    expect(result[2].text).toBe("world");
+    expect(result[2].number).toBe(123);
+    expect(result[2].bool).toBe(null);
+  });
+});

nodejs/__test__/connection.test.ts

@@ -203,3 +203,106 @@ describe("given a connection", () => {
     });
   });
 });
+
+describe("clone table functionality", () => {
+  let tmpDir: tmp.DirResult;
+  let db: Connection;
+  beforeEach(async () => {
+    tmpDir = tmp.dirSync({ unsafeCleanup: true });
+    db = await connect(tmpDir.name);
+  });
+  afterEach(() => tmpDir.removeCallback());
+
+  it("should clone a table with latest version (default behavior)", async () => {
+    // Create source table with some data
+    const data = [
+      { id: 1, text: "hello", vector: [1.0, 2.0] },
+      { id: 2, text: "world", vector: [3.0, 4.0] },
+    ];
+    const sourceTable = await db.createTable("source", data);
+
+    // Add more data to create a new version
+    const moreData = [{ id: 3, text: "test", vector: [5.0, 6.0] }];
+    await sourceTable.add(moreData);
+
+    // Clone the table (should get latest version with 3 rows)
+    const sourceUri = `${tmpDir.name}/source.lance`;
+    const clonedTable = await db.cloneTable("cloned", sourceUri);
+
+    // Verify cloned table has all 3 rows
+    expect(await clonedTable.countRows()).toBe(3);
+    expect((await db.tableNames()).includes("cloned")).toBe(true);
+  });
+
+  it("should clone a table from a specific version", async () => {
+    // Create source table with initial data
+    const data = [
+      { id: 1, text: "hello", vector: [1.0, 2.0] },
+      { id: 2, text: "world", vector: [3.0, 4.0] },
+    ];
+    const sourceTable = await db.createTable("source", data);
+
+    // Get the initial version
+    const initialVersion = await sourceTable.version();
+
+    // Add more data to create a new version
+    const moreData = [{ id: 3, text: "test", vector: [5.0, 6.0] }];
+    await sourceTable.add(moreData);
+
+    // Verify source now has 3 rows
+    expect(await sourceTable.countRows()).toBe(3);
+
+    // Clone from the initial version (should have only 2 rows)
+    const sourceUri = `${tmpDir.name}/source.lance`;
+    const clonedTable = await db.cloneTable("cloned", sourceUri, {
+      sourceVersion: initialVersion,
+    });
+
+    // Verify cloned table has only the initial 2 rows
+    expect(await clonedTable.countRows()).toBe(2);
+  });
+
+  it("should clone a table from a tagged version", async () => {
+    // Create source table with initial data
+    const data = [
+      { id: 1, text: "hello", vector: [1.0, 2.0] },
+      { id: 2, text: "world", vector: [3.0, 4.0] },
+    ];
+    const sourceTable = await db.createTable("source", data);
+
+    // Create a tag for the current version
+    const tags = await sourceTable.tags();
+    await tags.create("v1.0", await sourceTable.version());
+
+    // Add more data after the tag
+    const moreData = [{ id: 3, text: "test", vector: [5.0, 6.0] }];
+    await sourceTable.add(moreData);
+
+    // Verify source now has 3 rows
+    expect(await sourceTable.countRows()).toBe(3);
+
+    // Clone from the tagged version (should have only 2 rows)
+    const sourceUri = `${tmpDir.name}/source.lance`;
+    const clonedTable = await db.cloneTable("cloned", sourceUri, {
+      sourceTag: "v1.0",
+    });
+
+    // Verify cloned table has only the tagged version's 2 rows
+    expect(await clonedTable.countRows()).toBe(2);
+  });
+
+  it("should fail when attempting deep clone", async () => {
+    // Create source table with some data
+    const data = [
+      { id: 1, text: "hello", vector: [1.0, 2.0] },
+      { id: 2, text: "world", vector: [3.0, 4.0] },
+    ];
+    await db.createTable("source", data);
+
+    // Try to create a deep clone (should fail)
+    const sourceUri = `${tmpDir.name}/source.lance`;
+    await expect(
+      db.cloneTable("cloned", sourceUri, { isShallow: false }),
+    ).rejects.toThrow("Deep clone is not yet implemented");
+  });
+});

nodejs/__test__/embedding.test.ts

@@ -256,6 +256,60 @@ describe("embedding functions", () => {
     expect(actual).toHaveProperty("text");
   });
 
+  it("should handle undefined vector field with embedding function correctly", async () => {
+    @register("undefined_test")
+    class MockEmbeddingFunction extends EmbeddingFunction<string> {
+      ndims() {
+        return 3;
+      }
+      embeddingDataType(): Float {
+        return new Float32();
+      }
+      async computeQueryEmbeddings(_data: string) {
+        return [1, 2, 3];
+      }
+      async computeSourceEmbeddings(data: string[]) {
+        return Array.from({ length: data.length }).fill([
+          1, 2, 3,
+        ]) as number[][];
+      }
+    }
+    const func = getRegistry()
+      .get<MockEmbeddingFunction>("undefined_test")!
+      .create();
+    const schema = new Schema([
+      new Field("text", new Utf8(), true),
+      new Field(
+        "vector",
+        new FixedSizeList(3, new Field("item", new Float32(), true)),
+        true,
+      ),
+    ]);
+
+    const db = await connect(tmpDir.name);
+    const table = await db.createEmptyTable("test_undefined", schema, {
+      embeddingFunction: {
+        function: func,
+        sourceColumn: "text",
+        vectorColumn: "vector",
+      },
+    });
+
+    // Test that undefined, null, and omitted vector fields all work
+    await table.add([{ text: "test1", vector: undefined }]);
+    await table.add([{ text: "test2", vector: null }]);
+    await table.add([{ text: "test3" }]);
+
+    const rows = await table.query().toArray();
+    expect(rows.length).toBe(3);
+
+    // All rows should have vectors computed by the embedding function
+    for (const row of rows) {
+      expect(row.vector).toBeDefined();
+      expect(JSON.parse(JSON.stringify(row.vector))).toEqual([1, 2, 3]);
+    }
+  });
+
   test.each([new Float16(), new Float32(), new Float64()])(
     "should be able to provide manual embeddings with multiple float datatype",
     async (floatType) => {

nodejs/__test__/remote.test.ts

@@ -3,7 +3,50 @@
 
 import * as http from "http";
 import { RequestListener } from "http";
-import { Connection, ConnectionOptions, connect } from "../lancedb";
+import {
+  ClientConfig,
+  Connection,
+  ConnectionOptions,
+  NativeJsHeaderProvider,
+  TlsConfig,
+  connect,
+} from "../lancedb";
+import {
+  HeaderProvider,
+  OAuthHeaderProvider,
+  StaticHeaderProvider,
+} from "../lancedb/header";
+
+// Test-only header providers
+class CustomProvider extends HeaderProvider {
+  getHeaders(): Record<string, string> {
+    return { "X-Custom": "custom-value" };
+  }
+}
+
+class ErrorProvider extends HeaderProvider {
+  private errorMessage: string;
+  public callCount: number = 0;
+
+  constructor(errorMessage: string = "Test error") {
+    super();
+    this.errorMessage = errorMessage;
+  }
+
+  getHeaders(): Record<string, string> {
+    this.callCount++;
+    throw new Error(this.errorMessage);
+  }
+}
+
+class ConcurrentProvider extends HeaderProvider {
+  private counter: number = 0;
+
+  getHeaders(): Record<string, string> {
+    this.counter++;
+    return { "X-Request-Id": String(this.counter) };
+  }
+}
 
 async function withMockDatabase(
   listener: RequestListener,
@@ -148,4 +191,431 @@ describe("remote connection", () => {
       },
     );
   });
+
+  describe("TlsConfig", () => {
+    it("should create TlsConfig with all fields", () => {
+      const tlsConfig: TlsConfig = {
+        certFile: "/path/to/cert.pem",
+        keyFile: "/path/to/key.pem",
+        sslCaCert: "/path/to/ca.pem",
+        assertHostname: false,
+      };
+
+      expect(tlsConfig.certFile).toBe("/path/to/cert.pem");
+      expect(tlsConfig.keyFile).toBe("/path/to/key.pem");
+      expect(tlsConfig.sslCaCert).toBe("/path/to/ca.pem");
+      expect(tlsConfig.assertHostname).toBe(false);
+    });
+
+    it("should create TlsConfig with partial fields", () => {
+      const tlsConfig: TlsConfig = {
+        certFile: "/path/to/cert.pem",
+        keyFile: "/path/to/key.pem",
+      };
+
+      expect(tlsConfig.certFile).toBe("/path/to/cert.pem");
+      expect(tlsConfig.keyFile).toBe("/path/to/key.pem");
+      expect(tlsConfig.sslCaCert).toBeUndefined();
+      expect(tlsConfig.assertHostname).toBeUndefined();
+    });
+
+    it("should create ClientConfig with TlsConfig", () => {
+      const tlsConfig: TlsConfig = {
+        certFile: "/path/to/cert.pem",
+        keyFile: "/path/to/key.pem",
+        sslCaCert: "/path/to/ca.pem",
+        assertHostname: true,
+      };
+
+      const clientConfig: ClientConfig = {
+        userAgent: "test-agent",
+        tlsConfig: tlsConfig,
+      };
+
+      expect(clientConfig.userAgent).toBe("test-agent");
+      expect(clientConfig.tlsConfig).toBeDefined();
+      expect(clientConfig.tlsConfig?.certFile).toBe("/path/to/cert.pem");
+      expect(clientConfig.tlsConfig?.keyFile).toBe("/path/to/key.pem");
+      expect(clientConfig.tlsConfig?.sslCaCert).toBe("/path/to/ca.pem");
+      expect(clientConfig.tlsConfig?.assertHostname).toBe(true);
+    });
+
+    it("should handle empty TlsConfig", () => {
+      const tlsConfig: TlsConfig = {};
+
+      expect(tlsConfig.certFile).toBeUndefined();
+      expect(tlsConfig.keyFile).toBeUndefined();
+      expect(tlsConfig.sslCaCert).toBeUndefined();
+      expect(tlsConfig.assertHostname).toBeUndefined();
+    });
+
+    it("should accept TlsConfig in connection options", () => {
+      const tlsConfig: TlsConfig = {
+        certFile: "/path/to/cert.pem",
+        keyFile: "/path/to/key.pem",
+        sslCaCert: "/path/to/ca.pem",
+        assertHostname: false,
+      };
+
+      // Just verify that the ClientConfig accepts the TlsConfig
+      const clientConfig: ClientConfig = {
+        tlsConfig: tlsConfig,
+      };
+
+      const connectionOptions: ConnectionOptions = {
+        apiKey: "fake",
+        clientConfig: clientConfig,
+      };
+
+      // Verify the configuration structure is correct
+      expect(connectionOptions.clientConfig).toBeDefined();
+      expect(connectionOptions.clientConfig?.tlsConfig).toBeDefined();
+      expect(connectionOptions.clientConfig?.tlsConfig?.certFile).toBe(
+        "/path/to/cert.pem",
+      );
+    });
+  });
+
+  describe("header providers", () => {
+    it("should work with StaticHeaderProvider", async () => {
+      const provider = new StaticHeaderProvider({
+        authorization: "Bearer test-token",
+        "X-Custom": "value",
+      });
+
+      const headers = provider.getHeaders();
+      expect(headers).toEqual({
+        authorization: "Bearer test-token",
+        "X-Custom": "value",
+      });
+
+      // Test that it returns a copy
+      headers["X-Modified"] = "modified";
+      const headers2 = provider.getHeaders();
+      expect(headers2).not.toHaveProperty("X-Modified");
+    });
+
+    it("should pass headers from StaticHeaderProvider to requests", async () => {
+      const provider = new StaticHeaderProvider({
+        "X-Custom-Auth": "secret-token",
+        "X-Request-Source": "test-suite",
+      });
+
+      await withMockDatabase(
+        (req, res) => {
+          expect(req.headers["x-custom-auth"]).toEqual("secret-token");
+          expect(req.headers["x-request-source"]).toEqual("test-suite");
+
+          const body = JSON.stringify({ tables: [] });
+          res.writeHead(200, { "Content-Type": "application/json" }).end(body);
+        },
+        async () => {
+          // Use actual header provider mechanism instead of extraHeaders
+          const conn = await connect(
+            "db://dev",
+            {
+              apiKey: "fake",
+              hostOverride: "http://localhost:8000",
+            },
+            undefined, // session
+            provider, // headerProvider
+          );
+
+          const tableNames = await conn.tableNames();
+          expect(tableNames).toEqual([]);
+        },
+      );
+    });
+
+    it("should work with CustomProvider", () => {
+      const provider = new CustomProvider();
+      const headers = provider.getHeaders();
+      expect(headers).toEqual({ "X-Custom": "custom-value" });
+    });
+
+    it("should handle ErrorProvider errors", () => {
+      const provider = new ErrorProvider("Authentication failed");
+
+      expect(() => provider.getHeaders()).toThrow("Authentication failed");
+      expect(provider.callCount).toBe(1);
+
+      // Test that error is thrown each time
+      expect(() => provider.getHeaders()).toThrow("Authentication failed");
+      expect(provider.callCount).toBe(2);
+    });
+
+    it("should work with ConcurrentProvider", () => {
+      const provider = new ConcurrentProvider();
+
+      const headers1 = provider.getHeaders();
+      const headers2 = provider.getHeaders();
+      const headers3 = provider.getHeaders();
+
+      expect(headers1).toEqual({ "X-Request-Id": "1" });
+      expect(headers2).toEqual({ "X-Request-Id": "2" });
+      expect(headers3).toEqual({ "X-Request-Id": "3" });
+    });
+
+    describe("OAuthHeaderProvider", () => {
+      it("should initialize correctly", () => {
+        const fetcher = () => ({
+          accessToken: "token123",
+          expiresIn: 3600,
+        });
+
+        const provider = new OAuthHeaderProvider(fetcher);
+        expect(provider).toBeInstanceOf(HeaderProvider);
+      });
+
+      it("should fetch token on first use", async () => {
+        let callCount = 0;
+        const fetcher = () => {
+          callCount++;
+          return {
+            accessToken: "token123",
+            expiresIn: 3600,
+          };
+        };
+
+        const provider = new OAuthHeaderProvider(fetcher);
+
+        // Need to manually refresh first due to sync limitation
+        await provider.refreshToken();
+
+        const headers = provider.getHeaders();
+        expect(headers).toEqual({ authorization: "Bearer token123" });
+        expect(callCount).toBe(1);
+
+        // Second call should not fetch again
+        const headers2 = provider.getHeaders();
+        expect(headers2).toEqual({ authorization: "Bearer token123" });
+        expect(callCount).toBe(1);
+      });
+
+      it("should handle tokens without expiry", async () => {
+        const fetcher = () => ({
+          accessToken: "permanent_token",
+        });
+
+        const provider = new OAuthHeaderProvider(fetcher);
+        await provider.refreshToken();
+
+        const headers = provider.getHeaders();
+        expect(headers).toEqual({ authorization: "Bearer permanent_token" });
+      });
+
+      it("should throw error when access_token is missing", async () => {
+        const fetcher = () =>
+          ({
+            expiresIn: 3600,
+          }) as { accessToken?: string; expiresIn?: number };
+
+        const provider = new OAuthHeaderProvider(
+          fetcher as () => {
+            accessToken: string;
+            expiresIn?: number;
+          },
+        );
+
+        await expect(provider.refreshToken()).rejects.toThrow(
+          "Token fetcher did not return 'accessToken'",
+        );
+      });
+
+      it("should handle async token fetchers", async () => {
+        const fetcher = async () => {
+          // Simulate async operation
+          await new Promise((resolve) => setTimeout(resolve, 10));
+          return {
+            accessToken: "async_token",
+            expiresIn: 3600,
+          };
+        };
+
+        const provider = new OAuthHeaderProvider(fetcher);
+        await provider.refreshToken();
+
+        const headers = provider.getHeaders();
+        expect(headers).toEqual({ authorization: "Bearer async_token" });
+      });
+    });
+
+    it("should merge header provider headers with extra headers", async () => {
+      const provider = new StaticHeaderProvider({
+        "X-From-Provider": "provider-value",
+      });
+
+      await withMockDatabase(
+        (req, res) => {
+          expect(req.headers["x-from-provider"]).toEqual("provider-value");
+          expect(req.headers["x-extra-header"]).toEqual("extra-value");
+
+          const body = JSON.stringify({ tables: [] });
+          res.writeHead(200, { "Content-Type": "application/json" }).end(body);
+        },
+        async () => {
+          // Use header provider with additional extraHeaders
+          const conn = await connect(
+            "db://dev",
+            {
+              apiKey: "fake",
+              hostOverride: "http://localhost:8000",
+              clientConfig: {
+                extraHeaders: {
+                  "X-Extra-Header": "extra-value",
+                },
+              },
+            },
+            undefined, // session
+            provider, // headerProvider
+          );
+
+          const tableNames = await conn.tableNames();
+          expect(tableNames).toEqual([]);
+        },
+      );
+    });
+  });
+
+  describe("header provider integration", () => {
+    it("should work with TypeScript StaticHeaderProvider", async () => {
+      let requestCount = 0;
+
+      await withMockDatabase(
+        (req, res) => {
+          requestCount++;
+
+          // Check headers are present on each request
+          expect(req.headers["authorization"]).toEqual("Bearer test-token-123");
+          expect(req.headers["x-custom"]).toEqual("custom-value");
+
+          // Return different responses based on the endpoint
+          if (req.url === "/v1/table/test_table/describe/") {
+            const body = JSON.stringify({
+              name: "test_table",
+              schema: { fields: [] },
+            });
+            res
+              .writeHead(200, { "Content-Type": "application/json" })
+              .end(body);
+          } else {
+            const body = JSON.stringify({ tables: ["test_table"] });
+            res
+              .writeHead(200, { "Content-Type": "application/json" })
+              .end(body);
+          }
+        },
+        async () => {
+          // Create provider with static headers
+          const provider = new StaticHeaderProvider({
+            authorization: "Bearer test-token-123",
+            "X-Custom": "custom-value",
+          });
+
+          // Connect with the provider
+          const conn = await connect(
+            "db://dev",
+            {
+              apiKey: "fake",
+              hostOverride: "http://localhost:8000",
+            },
+            undefined, // session
+            provider, // headerProvider
+          );
+
+          // Make multiple requests to verify headers are sent each time
+          const tables1 = await conn.tableNames();
+          expect(tables1).toEqual(["test_table"]);
+
+          const tables2 = await conn.tableNames();
+          expect(tables2).toEqual(["test_table"]);
+
+          // Verify headers were sent with each request
+          expect(requestCount).toBeGreaterThanOrEqual(2);
+        },
+      );
+    });
+
+    it("should work with JavaScript function provider", async () => {
+      let requestId = 0;
+
+      await withMockDatabase(
+        (req, res) => {
+          // Check dynamic header is present
+          expect(req.headers["x-request-id"]).toBeDefined();
+          expect(req.headers["x-request-id"]).toMatch(/^req-\d+$/);
+
+          const body = JSON.stringify({ tables: [] });
+          res.writeHead(200, { "Content-Type": "application/json" }).end(body);
+        },
+        async () => {
+          // Create a JavaScript function that returns dynamic headers
+          const getHeaders = async () => {
+            requestId++;
+            return {
+              "X-Request-Id": `req-${requestId}`,
+              "X-Timestamp": new Date().toISOString(),
+            };
+          };
+
+          // Connect with the function directly
+          const conn = await connect(
+            "db://dev",
+            {
+              apiKey: "fake",
+              hostOverride: "http://localhost:8000",
+            },
+            undefined, // session
+            getHeaders, // headerProvider
+          );
+
+          // Make requests - each should have different headers
+          const tables = await conn.tableNames();
+          expect(tables).toEqual([]);
+        },
+      );
+    });
+
+    it("should support OAuth-like token refresh pattern", async () => {
+      let tokenVersion = 0;
+
+      await withMockDatabase(
+        (req, res) => {
+          // Verify authorization header
+          const authHeader = req.headers["authorization"];
+          expect(authHeader).toBeDefined();
+          expect(authHeader).toMatch(/^Bearer token-v\d+$/);
+
+          const body = JSON.stringify({ tables: [] });
+          res.writeHead(200, { "Content-Type": "application/json" }).end(body);
+        },
+        async () => {
+          // Simulate OAuth token fetcher
+          const fetchToken = async () => {
+            tokenVersion++;
+            return {
+              authorization: `Bearer token-v${tokenVersion}`,
+            };
+          };
+
+          // Connect with the function directly
+          const conn = await connect(
+            "db://dev",
+            {
+              apiKey: "fake",
+              hostOverride: "http://localhost:8000",
+            },
+            undefined, // session
+            fetchToken, // headerProvider
+          );
+
+          // Each request will fetch a new token
+          await conn.tableNames();
+
+          // Token should be different on next request
+          await conn.tableNames();
+        },
+      );
+    });
+  });
 });

nodejs/__test__/table.test.ts

@@ -287,6 +287,12 @@ describe.each([arrow15, arrow16, arrow17, arrow18])(
       expect(res2[1].id).toEqual(data2.id);
     });
 
+    it("should support take queries", async () => {
+      await table.add([{ id: 1 }, { id: 2 }, { id: 3 }]);
+      const res = await table.takeOffsets([1, 2]).toArrow();
+      expect(res.getChild("id")?.toJSON()).toEqual([2, 3]);
+    });
+
     it("should return the table as an instance of an arrow table", async () => {
       const arrowTbl = await table.toArrow();
       expect(arrowTbl).toBeInstanceOf(ArrowTable);
@@ -557,7 +563,7 @@ describe("When creating an index", () => {
 
     // test offset
     rst = await tbl.query().limit(2).offset(1).nearestTo(queryVec).toArrow();
-    expect(rst.numRows).toBe(1);
+    expect(rst.numRows).toBe(2);
 
     // test nprobes
     rst = await tbl.query().nearestTo(queryVec).limit(2).nprobes(50).toArrow();
@@ -696,7 +702,7 @@ describe("When creating an index", () => {
 
     // test offset
     rst = await tbl.query().limit(2).offset(1).nearestTo(queryVec).toArrow();
-    expect(rst.numRows).toBe(1);
+    expect(rst.numRows).toBe(2);
 
     // test ef
     rst = await tbl.query().limit(2).nearestTo(queryVec).ef(100).toArrow();
@@ -851,6 +857,40 @@ describe("When creating an index", () => {
     expect(stats).toBeUndefined();
   });
 
+  test("should support name and train parameters", async () => {
+    // Test with custom name
+    await tbl.createIndex("vec", {
+      config: Index.ivfPq({ numPartitions: 4 }),
+      name: "my_custom_vector_index",
+    });
+
+    const indices = await tbl.listIndices();
+    expect(indices).toHaveLength(1);
+    expect(indices[0].name).toBe("my_custom_vector_index");
+
+    // Test scalar index with train=false
+    await tbl.createIndex("id", {
+      config: Index.btree(),
+      name: "btree_empty",
+      train: false,
+    });
+
+    const allIndices = await tbl.listIndices();
+    expect(allIndices).toHaveLength(2);
+    expect(allIndices.some((idx) => idx.name === "btree_empty")).toBe(true);
+
+    // Test with both name and train=true (use tags column)
+    await tbl.createIndex("tags", {
+      config: Index.labelList(),
+      name: "tags_trained",
+      train: true,
+    });
+
+    const finalIndices = await tbl.listIndices();
+    expect(finalIndices).toHaveLength(3);
+    expect(finalIndices.some((idx) => idx.name === "tags_trained")).toBe(true);
+  });
+
   test("create ivf_flat with binary vectors", async () => {
     const db = await connect(tmpDir.name);
     const binarySchema = new Schema([

nodejs/examples/ann_indexes.test.ts

@@ -12,7 +12,7 @@ test("ann index examples", async () => {
     // --8<-- [start:ingest]
     const db = await lancedb.connect(databaseDir);
 
-    const data = Array.from({ length: 5_000 }, (_, i) => ({
+    const data = Array.from({ length: 1_000 }, (_, i) => ({
       vector: Array(128).fill(i),
       id: `${i}`,
       content: "",
@@ -24,8 +24,8 @@ test("ann index examples", async () => {
     });
     await table.createIndex("vector", {
       config: lancedb.Index.ivfPq({
-        numPartitions: 10,
-        numSubVectors: 16,
+        numPartitions: 30,
+        numSubVectors: 8,
       }),
     });
     // --8<-- [end:ingest]

nodejs/lancedb/arrow.ts

@@ -512,7 +512,11 @@ function* rowPathsAndValues(
     if (isObject(value)) {
       yield* rowPathsAndValues(value, [...basePath, key]);
     } else {
-      yield [[...basePath, key], value];
+      // Skip undefined values - they should be treated the same as missing fields
+      // for embedding function purposes
+      if (value !== undefined) {
+        yield [[...basePath, key], value];
+      }
     }
   }
 }
@@ -701,7 +705,7 @@ function transposeData(
       }
       return current;
     });
-    return makeVector(values, field.type);
+    return makeVector(values, field.type, undefined, field.nullable);
   }
 }
 
@@ -748,9 +752,30 @@ function makeVector(
   values: unknown[],
   type?: DataType,
   stringAsDictionary?: boolean,
+  nullable?: boolean,
   // biome-ignore lint/suspicious/noExplicitAny: skip
 ): Vector<any> {
   if (type !== undefined) {
+    // Convert undefined values to null for nullable fields
+    if (nullable) {
+      values = values.map((v) => (v === undefined ? null : v));
+    }
+
+    // workaround for: https://github.com/apache/arrow-js/issues/68
+    if (DataType.isBool(type)) {
+      const hasNonNullValue = values.some((v) => v !== null && v !== undefined);
+      if (!hasNonNullValue) {
+        const nullBitmap = new Uint8Array(Math.ceil(values.length / 8));
+        const data = makeData({
+          type: type,
+          length: values.length,
+          nullCount: values.length,
+          nullBitmap,
+        });
+        return arrowMakeVector(data);
+      }
+    }
+
     // No need for inference, let Arrow create it
     if (type instanceof Int) {
       if (DataType.isInt(type) && type.bitWidth === 64) {
@@ -875,7 +900,12 @@ async function applyEmbeddingsFromMetadata(
   for (const field of schema.fields) {
     if (!(field.name in columns)) {
       const nullValues = new Array(table.numRows).fill(null);
-      columns[field.name] = makeVector(nullValues, field.type);
+      columns[field.name] = makeVector(
+        nullValues,
+        field.type,
+        undefined,
+        field.nullable,
+      );
     }
   }
 
@@ -939,7 +969,12 @@ async function applyEmbeddings<T>(
     } else if (schema != null) {
       const destField = schema.fields.find((f) => f.name === destColumn);
       if (destField != null) {
-        newColumns[destColumn] = makeVector([], destField.type);
+        newColumns[destColumn] = makeVector(
+          [],
+          destField.type,
+          undefined,
+          destField.nullable,
+        );
       } else {
         throw new Error(
           `Attempt to apply embeddings to an empty table failed because schema was missing embedding column '${destColumn}'`,

nodejs/lancedb/connection.ts

@@ -159,17 +159,33 @@ export abstract class Connection {
    *
    * Tables will be returned in lexicographical order.
    * @param {Partial<TableNamesOptions>} options - options to control the
-   * paging / start point
+   * paging / start point (backwards compatibility)
    *
    */
   abstract tableNames(options?: Partial<TableNamesOptions>): Promise<string[]>;
+  /**
+   * List all the table names in this database.
+   *
+   * Tables will be returned in lexicographical order.
+   * @param {string[]} namespace - The namespace to list tables from (defaults to root namespace)
+   * @param {Partial<TableNamesOptions>} options - options to control the
+   * paging / start point
+   *
+   */
+  abstract tableNames(
+    namespace?: string[],
+    options?: Partial<TableNamesOptions>,
+  ): Promise<string[]>;
 
   /**
    * Open a table in the database.
    * @param {string} name - The name of the table
+   * @param {string[]} namespace - The namespace of the table (defaults to root namespace)
+   * @param {Partial<OpenTableOptions>} options - Additional options
    */
   abstract openTable(
     name: string,
+    namespace?: string[],
     options?: Partial<OpenTableOptions>,
   ): Promise<Table>;
 
@@ -178,6 +194,7 @@ export abstract class Connection {
    * @param {object} options - The options object.
    * @param {string} options.name - The name of the table.
    * @param {Data} options.data - Non-empty Array of Records to be inserted into the table
+   * @param {string[]} namespace - The namespace to create the table in (defaults to root namespace)
    *
    */
   abstract createTable(
@@ -185,40 +202,99 @@ export abstract class Connection {
       name: string;
       data: Data;
     } & Partial<CreateTableOptions>,
+    namespace?: string[],
   ): Promise<Table>;
   /**
    * Creates a new Table and initialize it with new data.
    * @param {string} name - The name of the table.
    * @param {Record<string, unknown>[] | TableLike} data - Non-empty Array of Records
    * to be inserted into the table
+   * @param {Partial<CreateTableOptions>} options - Additional options (backwards compatibility)
    */
   abstract createTable(
     name: string,
     data: Record<string, unknown>[] | TableLike,
     options?: Partial<CreateTableOptions>,
   ): Promise<Table>;
+  /**
+   * Creates a new Table and initialize it with new data.
+   * @param {string} name - The name of the table.
+   * @param {Record<string, unknown>[] | TableLike} data - Non-empty Array of Records
+   * to be inserted into the table
+   * @param {string[]} namespace - The namespace to create the table in (defaults to root namespace)
+   * @param {Partial<CreateTableOptions>} options - Additional options
+   */
+  abstract createTable(
+    name: string,
+    data: Record<string, unknown>[] | TableLike,
+    namespace?: string[],
+    options?: Partial<CreateTableOptions>,
+  ): Promise<Table>;
 
   /**
    * Creates a new empty Table
    * @param {string} name - The name of the table.
    * @param {Schema} schema - The schema of the table
+   * @param {Partial<CreateTableOptions>} options - Additional options (backwards compatibility)
    */
   abstract createEmptyTable(
     name: string,
     schema: import("./arrow").SchemaLike,
     options?: Partial<CreateTableOptions>,
   ): Promise<Table>;
+  /**
+   * Creates a new empty Table
+   * @param {string} name - The name of the table.
+   * @param {Schema} schema - The schema of the table
+   * @param {string[]} namespace - The namespace to create the table in (defaults to root namespace)
+   * @param {Partial<CreateTableOptions>} options - Additional options
+   */
+  abstract createEmptyTable(
+    name: string,
+    schema: import("./arrow").SchemaLike,
+    namespace?: string[],
+    options?: Partial<CreateTableOptions>,
+  ): Promise<Table>;
 
   /**
    * Drop an existing table.
    * @param {string} name The name of the table to drop.
+   * @param {string[]} namespace The namespace of the table (defaults to root namespace).
    */
-  abstract dropTable(name: string): Promise<void>;
+  abstract dropTable(name: string, namespace?: string[]): Promise<void>;
 
   /**
    * Drop all tables in the database.
+   * @param {string[]} namespace The namespace to drop tables from (defaults to root namespace).
    */
-  abstract dropAllTables(): Promise<void>;
+  abstract dropAllTables(namespace?: string[]): Promise<void>;
+
+  /**
+   * Clone a table from a source table.
+   *
+   * A shallow clone creates a new table that shares the underlying data files
+   * with the source table but has its own independent manifest. This allows
+   * both the source and cloned tables to evolve independently while initially
+   * sharing the same data, deletion, and index files.
+   *
+   * @param {string} targetTableName - The name of the target table to create.
+   * @param {string} sourceUri - The URI of the source table to clone from.
+   * @param {object} options - Clone options.
+   * @param {string[]} options.targetNamespace - The namespace for the target table (defaults to root namespace).
+   * @param {number} options.sourceVersion - The version of the source table to clone.
+   * @param {string} options.sourceTag - The tag of the source table to clone.
+   * @param {boolean} options.isShallow - Whether to perform a shallow clone (defaults to true).
+   */
+  abstract cloneTable(
+    targetTableName: string,
+    sourceUri: string,
+    options?: {
+      targetNamespace?: string[];
+      sourceVersion?: number;
+      sourceTag?: string;
+      isShallow?: boolean;
+    },
+  ): Promise<Table>;
 }
 
 /** @hideconstructor */
@@ -243,16 +319,39 @@ export class LocalConnection extends Connection {
     return this.inner.display();
   }
 
-  async tableNames(options?: Partial<TableNamesOptions>): Promise<string[]> {
-    return this.inner.tableNames(options?.startAfter, options?.limit);
+  async tableNames(
+    namespaceOrOptions?: string[] | Partial<TableNamesOptions>,
+    options?: Partial<TableNamesOptions>,
+  ): Promise<string[]> {
+    // Detect if first argument is namespace array or options object
+    let namespace: string[] | undefined;
+    let tableNamesOptions: Partial<TableNamesOptions> | undefined;
+
+    if (Array.isArray(namespaceOrOptions)) {
+      // First argument is namespace array
+      namespace = namespaceOrOptions;
+      tableNamesOptions = options;
+    } else {
+      // First argument is options object (backwards compatibility)
+      namespace = undefined;
+      tableNamesOptions = namespaceOrOptions;
+    }
+
+    return this.inner.tableNames(
+      namespace ?? [],
+      tableNamesOptions?.startAfter,
+      tableNamesOptions?.limit,
+    );
   }
 
   async openTable(
     name: string,
+    namespace?: string[],
     options?: Partial<OpenTableOptions>,
   ): Promise<Table> {
     const innerTable = await this.inner.openTable(
       name,
+      namespace ?? [],
       cleanseStorageOptions(options?.storageOptions),
       options?.indexCacheSize,
     );
@@ -260,6 +359,28 @@ export class LocalConnection extends Connection {
     return new LocalTable(innerTable);
   }
 
+  async cloneTable(
+    targetTableName: string,
+    sourceUri: string,
+    options?: {
+      targetNamespace?: string[];
+      sourceVersion?: number;
+      sourceTag?: string;
+      isShallow?: boolean;
+    },
+  ): Promise<Table> {
+    const innerTable = await this.inner.cloneTable(
+      targetTableName,
+      sourceUri,
+      options?.targetNamespace ?? [],
+      options?.sourceVersion ?? null,
+      options?.sourceTag ?? null,
+      options?.isShallow ?? true,
+    );
+
+    return new LocalTable(innerTable);
+  }
+
   private getStorageOptions(
     options?: Partial<CreateTableOptions>,
   ): Record<string, string> | undefined {
@@ -286,14 +407,44 @@ export class LocalConnection extends Connection {
     nameOrOptions:
       | string
       | ({ name: string; data: Data } & Partial<CreateTableOptions>),
-    data?: Record<string, unknown>[] | TableLike,
+    dataOrNamespace?: Record<string, unknown>[] | TableLike | string[],
+    namespaceOrOptions?: string[] | Partial<CreateTableOptions>,
     options?: Partial<CreateTableOptions>,
   ): Promise<Table> {
     if (typeof nameOrOptions !== "string" && "name" in nameOrOptions) {
-      const { name, data, ...options } = nameOrOptions;
-
-      return this.createTable(name, data, options);
+      // First overload: createTable(options, namespace?)
+      const { name, data, ...createOptions } = nameOrOptions;
+      const namespace = dataOrNamespace as string[] | undefined;
+      return this._createTableImpl(name, data, namespace, createOptions);
     }
+
+    // Second overload: createTable(name, data, namespace?, options?)
+    const name = nameOrOptions;
+    const data = dataOrNamespace as Record<string, unknown>[] | TableLike;
+
+    // Detect if third argument is namespace array or options object
+    let namespace: string[] | undefined;
+    let createOptions: Partial<CreateTableOptions> | undefined;
+
+    if (Array.isArray(namespaceOrOptions)) {
+      // Third argument is namespace array
+      namespace = namespaceOrOptions;
+      createOptions = options;
+    } else {
+      // Third argument is options object (backwards compatibility)
+      namespace = undefined;
+      createOptions = namespaceOrOptions;
+    }
+
+    return this._createTableImpl(name, data, namespace, createOptions);
+  }
+
+  private async _createTableImpl(
+    name: string,
+    data: Data,
+    namespace?: string[],
+    options?: Partial<CreateTableOptions>,
+  ): Promise<Table> {
     if (data === undefined) {
       throw new Error("data is required");
     }
@@ -302,9 +453,10 @@ export class LocalConnection extends Connection {
     const storageOptions = this.getStorageOptions(options);
 
     const innerTable = await this.inner.createTable(
-      nameOrOptions,
+      name,
       buf,
       mode,
+      namespace ?? [],
       storageOptions,
     );
 
@@ -314,39 +466,55 @@ export class LocalConnection extends Connection {
   async createEmptyTable(
     name: string,
     schema: import("./arrow").SchemaLike,
+    namespaceOrOptions?: string[] | Partial<CreateTableOptions>,
     options?: Partial<CreateTableOptions>,
   ): Promise<Table> {
-    let mode: string = options?.mode ?? "create";
-    const existOk = options?.existOk ?? false;
+    // Detect if third argument is namespace array or options object
+    let namespace: string[] | undefined;
+    let createOptions: Partial<CreateTableOptions> | undefined;
+
+    if (Array.isArray(namespaceOrOptions)) {
+      // Third argument is namespace array
+      namespace = namespaceOrOptions;
+      createOptions = options;
+    } else {
+      // Third argument is options object (backwards compatibility)
+      namespace = undefined;
+      createOptions = namespaceOrOptions;
+    }
+
+    let mode: string = createOptions?.mode ?? "create";
+    const existOk = createOptions?.existOk ?? false;
 
     if (mode === "create" && existOk) {
       mode = "exist_ok";
     }
     let metadata: Map<string, string> | undefined = undefined;
-    if (options?.embeddingFunction !== undefined) {
-      const embeddingFunction = options.embeddingFunction;
+    if (createOptions?.embeddingFunction !== undefined) {
+      const embeddingFunction = createOptions.embeddingFunction;
       const registry = getRegistry();
       metadata = registry.getTableMetadata([embeddingFunction]);
     }
 
-    const storageOptions = this.getStorageOptions(options);
+    const storageOptions = this.getStorageOptions(createOptions);
     const table = makeEmptyTable(schema, metadata);
     const buf = await fromTableToBuffer(table);
     const innerTable = await this.inner.createEmptyTable(
       name,
       buf,
       mode,
+      namespace ?? [],
       storageOptions,
     );
     return new LocalTable(innerTable);
   }
 
-  async dropTable(name: string): Promise<void> {
-    return this.inner.dropTable(name);
+  async dropTable(name: string, namespace?: string[]): Promise<void> {
+    return this.inner.dropTable(name, namespace ?? []);
   }
 
-  async dropAllTables(): Promise<void> {
-    return this.inner.dropAllTables();
+  async dropAllTables(namespace?: string[]): Promise<void> {
+    return this.inner.dropAllTables(namespace ?? []);
   }
 }
 

nodejs/lancedb/header.ts

@@ -0,0 +1,253 @@
+// SPDX-License-Identifier: Apache-2.0
+// SPDX-FileCopyrightText: Copyright The LanceDB Authors
+
+/**
+ * Header providers for LanceDB remote connections.
+ *
+ * This module provides a flexible header management framework for LanceDB remote
+ * connections, allowing users to implement custom header strategies for
+ * authentication, request tracking, custom metadata, or any other header-based
+ * requirements.
+ *
+ * @module header
+ */
+
+/**
+ * Abstract base class for providing custom headers for each request.
+ *
+ * Users can implement this interface to provide dynamic headers for various purposes
+ * such as authentication (OAuth tokens, API keys), request tracking (correlation IDs),
+ * custom metadata, or any other header-based requirements. The provider is called
+ * before each request to ensure fresh header values are always used.
+ *
+ * @example
+ * Simple JWT token provider:
+ * ```typescript
+ * class JWTProvider extends HeaderProvider {
+ *   constructor(private token: string) {
+ *     super();
+ *   }
+ *
+ *   getHeaders(): Record<string, string> {
+ *     return { authorization: `Bearer ${this.token}` };
+ *   }
+ * }
+ * ```
+ *
+ * @example
+ * Provider with request tracking:
+ * ```typescript
+ * class RequestTrackingProvider extends HeaderProvider {
+ *   constructor(private sessionId: string) {
+ *     super();
+ *   }
+ *
+ *   getHeaders(): Record<string, string> {
+ *     return {
+ *       "X-Session-Id": this.sessionId,
+ *       "X-Request-Id": `req-${Date.now()}`
+ *     };
+ *   }
+ * }
+ * ```
+ */
+export abstract class HeaderProvider {
+  /**
+   * Get the latest headers to be added to requests.
+   *
+   * This method is called before each request to the remote LanceDB server.
+   * Implementations should return headers that will be merged with existing headers.
+   *
+   * @returns Dictionary of header names to values to add to the request.
+   * @throws If unable to fetch headers, the exception will be propagated and the request will fail.
+   */
+  abstract getHeaders(): Record<string, string>;
+}
+
+/**
+ * Example implementation: A simple header provider that returns static headers.
+ *
+ * This is an example implementation showing how to create a HeaderProvider
+ * for cases where headers don't change during the session.
+ *
+ * @example
+ * ```typescript
+ * const provider = new StaticHeaderProvider({
+ *   authorization: "Bearer my-token",
+ *   "X-Custom-Header": "custom-value"
+ * });
+ * const headers = provider.getHeaders();
+ * // Returns: {authorization: 'Bearer my-token', 'X-Custom-Header': 'custom-value'}
+ * ```
+ */
+export class StaticHeaderProvider extends HeaderProvider {
+  private _headers: Record<string, string>;
+
+  /**
+   * Initialize with static headers.
+   * @param headers - Headers to return for every request.
+   */
+  constructor(headers: Record<string, string>) {
+    super();
+    this._headers = { ...headers };
+  }
+
+  /**
+   * Return the static headers.
+   * @returns Copy of the static headers.
+   */
+  getHeaders(): Record<string, string> {
+    return { ...this._headers };
+  }
+}
+
+/**
+ * Token response from OAuth provider.
+ * @public
+ */
+export interface TokenResponse {
+  accessToken: string;
+  expiresIn?: number;
+}
+
+/**
+ * Example implementation: OAuth token provider with automatic refresh.
+ *
+ * This is an example implementation showing how to manage OAuth tokens
+ * with automatic refresh when they expire.
+ *
+ * @example
+ * ```typescript
+ * async function fetchToken(): Promise<TokenResponse> {
+ *   const response = await fetch("https://oauth.example.com/token", {
+ *     method: "POST",
+ *     body: JSON.stringify({
+ *       grant_type: "client_credentials",
+ *       client_id: "your-client-id",
+ *       client_secret: "your-client-secret"
+ *     }),
+ *     headers: { "Content-Type": "application/json" }
+ *   });
+ *   const data = await response.json();
+ *   return {
+ *     accessToken: data.access_token,
+ *     expiresIn: data.expires_in
+ *   };
+ * }
+ *
+ * const provider = new OAuthHeaderProvider(fetchToken);
+ * const headers = provider.getHeaders();
+ * // Returns: {"authorization": "Bearer <your-token>"}
+ * ```
+ */
+export class OAuthHeaderProvider extends HeaderProvider {
+  private _tokenFetcher: () => Promise<TokenResponse> | TokenResponse;
+  private _refreshBufferSeconds: number;
+  private _currentToken: string | null = null;
+  private _tokenExpiresAt: number | null = null;
+  private _refreshPromise: Promise<void> | null = null;
+
+  /**
+   * Initialize the OAuth provider.
+   * @param tokenFetcher - Function to fetch new tokens. Should return object with 'accessToken' and optionally 'expiresIn'.
+   * @param refreshBufferSeconds - Seconds before expiry to refresh token. Default 300 (5 minutes).
+   */
+  constructor(
+    tokenFetcher: () => Promise<TokenResponse> | TokenResponse,
+    refreshBufferSeconds: number = 300,
+  ) {
+    super();
+    this._tokenFetcher = tokenFetcher;
+    this._refreshBufferSeconds = refreshBufferSeconds;
+  }
+
+  /**
+   * Check if token needs refresh.
+   */
+  private _needsRefresh(): boolean {
+    if (this._currentToken === null) {
+      return true;
+    }
+
+    if (this._tokenExpiresAt === null) {
+      // No expiration info, assume token is valid
+      return false;
+    }
+
+    // Refresh if we're within the buffer time of expiration
+    const now = Date.now() / 1000;
+    return now >= this._tokenExpiresAt - this._refreshBufferSeconds;
+  }
+
+  /**
+   * Refresh the token if it's expired or close to expiring.
+   */
+  private async _refreshTokenIfNeeded(): Promise<void> {
+    if (!this._needsRefresh()) {
+      return;
+    }
+
+    // If refresh is already in progress, wait for it
+    if (this._refreshPromise) {
+      await this._refreshPromise;
+      return;
+    }
+
+    // Start refresh
+    this._refreshPromise = (async () => {
+      try {
+        const tokenData = await this._tokenFetcher();
+
+        this._currentToken = tokenData.accessToken;
+        if (!this._currentToken) {
+          throw new Error("Token fetcher did not return 'accessToken'");
+        }
+
+        // Set expiration if provided
+        if (tokenData.expiresIn) {
+          this._tokenExpiresAt = Date.now() / 1000 + tokenData.expiresIn;
+        } else {
+          // Token doesn't expire or expiration unknown
+          this._tokenExpiresAt = null;
+        }
+      } finally {
+        this._refreshPromise = null;
+      }
+    })();
+
+    await this._refreshPromise;
+  }
+
+  /**
+   * Get OAuth headers, refreshing token if needed.
+   * Note: This is synchronous for now as the Rust implementation expects sync.
+   * In a real implementation, this would need to handle async properly.
+   * @returns Headers with Bearer token authorization.
+   * @throws If unable to fetch or refresh token.
+   */
+  getHeaders(): Record<string, string> {
+    // For simplicity in this example, we assume the token is already fetched
+    // In a real implementation, this would need to handle the async nature properly
+    if (!this._currentToken && !this._refreshPromise) {
+      // Synchronously trigger refresh - this is a limitation of the current implementation
+      throw new Error(
+        "Token not initialized. Call refreshToken() first or use async initialization.",
+      );
+    }
+
+    if (!this._currentToken) {
+      throw new Error("Failed to obtain OAuth token");
+    }
+
+    return { authorization: `Bearer ${this._currentToken}` };
+  }
+
+  /**
+   * Manually refresh the token.
+   * Call this before using getHeaders() to ensure token is available.
+   */
+  async refreshToken(): Promise<void> {
+    this._currentToken = null; // Force refresh
+    await this._refreshTokenIfNeeded();
+  }
+}

nodejs/lancedb/index.ts

@@ -10,9 +10,15 @@ import {
 import {
   ConnectionOptions,
   Connection as LanceDbConnection,
+  JsHeaderProvider as NativeJsHeaderProvider,
   Session,
 } from "./native.js";
 
+import { HeaderProvider } from "./header";
+
+// Re-export native header provider for use with connectWithHeaderProvider
+export { JsHeaderProvider as NativeJsHeaderProvider } from "./native.js";
+
 export {
   AddColumnsSql,
   ConnectionOptions,
@@ -21,6 +27,7 @@ export {
   ClientConfig,
   TimeoutConfig,
   RetryConfig,
+  TlsConfig,
   OptimizeStats,
   CompactionStats,
   RemovalStats,
@@ -59,6 +66,7 @@ export {
   Query,
   QueryBase,
   VectorQuery,
+  TakeQuery,
   QueryExecutionOptions,
   FullTextSearchOptions,
   RecordBatchIterator,
@@ -92,6 +100,13 @@ export {
   ColumnAlteration,
 } from "./table";
 
+export {
+  HeaderProvider,
+  StaticHeaderProvider,
+  OAuthHeaderProvider,
+  TokenResponse,
+} from "./header";
+
 export { MergeInsertBuilder, WriteExecutionOptions } from "./merge";
 
 export * as embedding from "./embedding";
@@ -130,11 +145,27 @@ export { IntoSql, packBits } from "./util";
  *   {storageOptions: {timeout: "60s"}
  * });
  * ```
+ * @example
+ * Using with a header provider for per-request authentication:
+ * ```ts
+ * const provider = new StaticHeaderProvider({
+ *   "X-API-Key": "my-key"
+ * });
+ * const conn = await connectWithHeaderProvider(
+ *   "db://host:port",
+ *   options,
+ *   provider
+ * );
+ * ```
  */
 export async function connect(
   uri: string,
   options?: Partial<ConnectionOptions>,
   session?: Session,
+  headerProvider?:
+    | HeaderProvider
+    | (() => Record<string, string>)
+    | (() => Promise<Record<string, string>>),
 ): Promise<Connection>;
 /**
  * Connect to a LanceDB instance at the given URI.
@@ -168,18 +199,58 @@ export async function connect(
 ): Promise<Connection>;
 export async function connect(
   uriOrOptions: string | (Partial<ConnectionOptions> & { uri: string }),
-  options?: Partial<ConnectionOptions>,
+  optionsOrSession?: Partial<ConnectionOptions> | Session,
+  sessionOrHeaderProvider?:
+    | Session
+    | HeaderProvider
+    | (() => Record<string, string>)
+    | (() => Promise<Record<string, string>>),
+  headerProvider?:
+    | HeaderProvider
+    | (() => Record<string, string>)
+    | (() => Promise<Record<string, string>>),
 ): Promise<Connection> {
   let uri: string | undefined;
   let finalOptions: Partial<ConnectionOptions> = {};
+  let finalHeaderProvider:
+    | HeaderProvider
+    | (() => Record<string, string>)
+    | (() => Promise<Record<string, string>>)
+    | undefined;
 
   if (typeof uriOrOptions !== "string") {
+    // First overload: connect(options)
     const { uri: uri_, ...opts } = uriOrOptions;
     uri = uri_;
     finalOptions = opts;
   } else {
+    // Second overload: connect(uri, options?, session?, headerProvider?)
     uri = uriOrOptions;
-    finalOptions = options || {};
+
+    // Handle optionsOrSession parameter
+    if (optionsOrSession && "inner" in optionsOrSession) {
+      // Second param is session, so no options provided
+      finalOptions = {};
+    } else {
+      // Second param is options
+      finalOptions = (optionsOrSession as Partial<ConnectionOptions>) || {};
+    }
+
+    // Handle sessionOrHeaderProvider parameter
+    if (
+      sessionOrHeaderProvider &&
+      (typeof sessionOrHeaderProvider === "function" ||
+        "getHeaders" in sessionOrHeaderProvider)
+    ) {
+      // Third param is header provider
+      finalHeaderProvider = sessionOrHeaderProvider as
+        | HeaderProvider
+        | (() => Record<string, string>)
+        | (() => Promise<Record<string, string>>);
+    } else {
+      // Third param is session, header provider is fourth param
+      finalHeaderProvider = headerProvider;
+    }
   }
 
   if (!uri) {
@@ -190,6 +261,26 @@ export async function connect(
   (<ConnectionOptions>finalOptions).storageOptions = cleanseStorageOptions(
     (<ConnectionOptions>finalOptions).storageOptions,
   );
-  const nativeConn = await LanceDbConnection.new(uri, finalOptions);
+
+  // Create native header provider if one was provided
+  let nativeProvider: NativeJsHeaderProvider | undefined;
+  if (finalHeaderProvider) {
+    if (typeof finalHeaderProvider === "function") {
+      nativeProvider = new NativeJsHeaderProvider(finalHeaderProvider);
+    } else if (
+      finalHeaderProvider &&
+      typeof finalHeaderProvider.getHeaders === "function"
+    ) {
+      nativeProvider = new NativeJsHeaderProvider(async () =>
+        finalHeaderProvider.getHeaders(),
+      );
+    }
+  }
+
+  const nativeConn = await LanceDbConnection.new(
+    uri,
+    finalOptions,
+    nativeProvider,
+  );
   return new LocalConnection(nativeConn);
 }

nodejs/lancedb/indices.ts

@@ -700,5 +700,27 @@ export interface IndexOptions {
    */
   replace?: boolean;
 
+  /**
+   * Timeout in seconds to wait for index creation to complete.
+   *
+   * If not specified, the method will return immediately after starting the index creation.
+   */
   waitTimeoutSeconds?: number;
+
+  /**
+   * Optional custom name for the index.
+   *
+   * If not provided, a default name will be generated based on the column name.
+   */
+  name?: string;
+
+  /**
+   * Whether to train the index with existing data.
+   *
+   * If true (default), the index will be trained with existing data in the table.
+   * If false, the index will be created empty and populated as new data is added.
+   *
+   * Note: This option is only supported for scalar indices. Vector indices always train.
+   */
+  train?: boolean;
 }

nodejs/lancedb/query.ts

@@ -15,6 +15,7 @@ import {
   RecordBatchIterator as NativeBatchIterator,
   Query as NativeQuery,
   Table as NativeTable,
+  TakeQuery as NativeTakeQuery,
   VectorQuery as NativeVectorQuery,
 } from "./native";
 import { Reranker } from "./rerankers";
@@ -50,7 +51,7 @@ export class RecordBatchIterator implements AsyncIterator<RecordBatch> {
 /* eslint-enable */
 
 class RecordBatchIterable<
-  NativeQueryType extends NativeQuery | NativeVectorQuery,
+  NativeQueryType extends NativeQuery | NativeVectorQuery | NativeTakeQuery,
 > implements AsyncIterable<RecordBatch>
 {
   private inner: NativeQueryType;
@@ -107,8 +108,9 @@ export interface FullTextSearchOptions {
  *
  * @hideconstructor
  */
-export class QueryBase<NativeQueryType extends NativeQuery | NativeVectorQuery>
-  implements AsyncIterable<RecordBatch>
+export class QueryBase<
+  NativeQueryType extends NativeQuery | NativeVectorQuery | NativeTakeQuery,
+> implements AsyncIterable<RecordBatch>
 {
   /**
    * @hidden
@@ -133,56 +135,6 @@ export class QueryBase<NativeQueryType extends NativeQuery | NativeVectorQuery>
       fn(this.inner);
     }
   }
-  /**
-   * A filter statement to be applied to this query.
-   *
-   * The filter should be supplied as an SQL query string.  For example:
-   * @example
-   * x > 10
-   * y > 0 AND y < 100
-   * x > 5 OR y = 'test'
-   *
-   * Filtering performance can often be improved by creating a scalar index
-   * on the filter column(s).
-   */
-  where(predicate: string): this {
-    this.doCall((inner: NativeQueryType) => inner.onlyIf(predicate));
-    return this;
-  }
-  /**
-   * A filter statement to be applied to this query.
-   * @see where
-   * @deprecated Use `where` instead
-   */
-  filter(predicate: string): this {
-    return this.where(predicate);
-  }
-
-  fullTextSearch(
-    query: string | FullTextQuery,
-    options?: Partial<FullTextSearchOptions>,
-  ): this {
-    let columns: string[] | null = null;
-    if (options) {
-      if (typeof options.columns === "string") {
-        columns = [options.columns];
-      } else if (Array.isArray(options.columns)) {
-        columns = options.columns;
-      }
-    }
-
-    this.doCall((inner: NativeQueryType) => {
-      if (typeof query === "string") {
-        inner.fullTextSearch({
-          query: query,
-          columns: columns,
-        });
-      } else {
-        inner.fullTextSearch({ query: query.inner });
-      }
-    });
-    return this;
-  }
 
   /**
    * Return only the specified columns.
@@ -241,33 +193,6 @@ export class QueryBase<NativeQueryType extends NativeQuery | NativeVectorQuery>
     return this;
   }
 
-  /**
-   * Set the maximum number of results to return.
-   *
-   * By default, a plain search has no limit.  If this method is not
-   * called then every valid row from the table will be returned.
-   */
-  limit(limit: number): this {
-    this.doCall((inner: NativeQueryType) => inner.limit(limit));
-    return this;
-  }
-
-  offset(offset: number): this {
-    this.doCall((inner: NativeQueryType) => inner.offset(offset));
-    return this;
-  }
-
-  /**
-   * Skip searching un-indexed data. This can make search faster, but will miss
-   * any data that is not yet indexed.
-   *
-   * Use {@link Table#optimize} to index all un-indexed data.
-   */
-  fastSearch(): this {
-    this.doCall((inner: NativeQueryType) => inner.fastSearch());
-    return this;
-  }
-
   /**
    * Whether to return the row id in the results.
    *
@@ -403,6 +328,100 @@ export class QueryBase<NativeQueryType extends NativeQuery | NativeVectorQuery>
   }
 }
 
+export class StandardQueryBase<
+    NativeQueryType extends NativeQuery | NativeVectorQuery,
+  >
+  extends QueryBase<NativeQueryType>
+  implements ExecutableQuery
+{
+  constructor(inner: NativeQueryType | Promise<NativeQueryType>) {
+    super(inner);
+  }
+
+  /**
+   * A filter statement to be applied to this query.
+   *
+   * The filter should be supplied as an SQL query string.  For example:
+   * @example
+   * x > 10
+   * y > 0 AND y < 100
+   * x > 5 OR y = 'test'
+   *
+   * Filtering performance can often be improved by creating a scalar index
+   * on the filter column(s).
+   */
+  where(predicate: string): this {
+    this.doCall((inner: NativeQueryType) => inner.onlyIf(predicate));
+    return this;
+  }
+  /**
+   * A filter statement to be applied to this query.
+   * @see where
+   * @deprecated Use `where` instead
+   */
+  filter(predicate: string): this {
+    return this.where(predicate);
+  }
+
+  fullTextSearch(
+    query: string | FullTextQuery,
+    options?: Partial<FullTextSearchOptions>,
+  ): this {
+    let columns: string[] | null = null;
+    if (options) {
+      if (typeof options.columns === "string") {
+        columns = [options.columns];
+      } else if (Array.isArray(options.columns)) {
+        columns = options.columns;
+      }
+    }
+
+    this.doCall((inner: NativeQueryType) => {
+      if (typeof query === "string") {
+        inner.fullTextSearch({
+          query: query,
+          columns: columns,
+        });
+      } else {
+        inner.fullTextSearch({ query: query.inner });
+      }
+    });
+    return this;
+  }
+
+  /**
+   * Set the maximum number of results to return.
+   *
+   * By default, a plain search has no limit.  If this method is not
+   * called then every valid row from the table will be returned.
+   */
+  limit(limit: number): this {
+    this.doCall((inner: NativeQueryType) => inner.limit(limit));
+    return this;
+  }
+
+  /**
+   * Set the number of rows to skip before returning results.
+   *
+   * This is useful for pagination.
+   */
+  offset(offset: number): this {
+    this.doCall((inner: NativeQueryType) => inner.offset(offset));
+    return this;
+  }
+
+  /**
+   * Skip searching un-indexed data. This can make search faster, but will miss
+   * any data that is not yet indexed.
+   *
+   * Use {@link Table#optimize} to index all un-indexed data.
+   */
+  fastSearch(): this {
+    this.doCall((inner: NativeQueryType) => inner.fastSearch());
+    return this;
+  }
+}
+
 /**
  * An interface for a query that can be executed
  *
@@ -419,7 +438,7 @@ export interface ExecutableQuery {}
  *
  * @hideconstructor
  */
-export class VectorQuery extends QueryBase<NativeVectorQuery> {
+export class VectorQuery extends StandardQueryBase<NativeVectorQuery> {
   /**
    * @hidden
    */
@@ -679,13 +698,24 @@ export class VectorQuery extends QueryBase<NativeVectorQuery> {
   }
 }
 
+/**
+ * A query that returns a subset of the rows in the table.
+ *
+ * @hideconstructor
+ */
+export class TakeQuery extends QueryBase<NativeTakeQuery> {
+  constructor(inner: NativeTakeQuery) {
+    super(inner);
+  }
+}
+
 /** A builder for LanceDB queries.
  *
  * @see {@link Table#query}, {@link Table#search}
  *
  * @hideconstructor
  */
-export class Query extends QueryBase<NativeQuery> {
+export class Query extends StandardQueryBase<NativeQuery> {
   /**
    * @hidden
    */

nodejs/lancedb/table.ts

@@ -35,6 +35,7 @@ import {
 import {
   FullTextQuery,
   Query,
+  TakeQuery,
   VectorQuery,
   instanceOfFullTextQuery,
 } from "./query";
@@ -336,6 +337,20 @@ export abstract class Table {
    */
   abstract query(): Query;
 
+  /**
+   * Create a query that returns a subset of the rows in the table.
+   * @param offsets The offsets of the rows to return.
+   * @returns A builder that can be used to parameterize the query.
+   */
+  abstract takeOffsets(offsets: number[]): TakeQuery;
+
+  /**
+   * Create a query that returns a subset of the rows in the table.
+   * @param rowIds The row ids of the rows to return.
+   * @returns A builder that can be used to parameterize the query.
+   */
+  abstract takeRowIds(rowIds: number[]): TakeQuery;
+
   /**
    * Create a search query to find the nearest neighbors
    * of the given query
@@ -647,6 +662,8 @@ export class LocalTable extends Table {
       column,
       options?.replace,
       options?.waitTimeoutSeconds,
+      options?.name,
+      options?.train,
     );
   }
 
@@ -665,6 +682,14 @@ export class LocalTable extends Table {
     await this.inner.waitForIndex(indexNames, timeoutSeconds);
   }
 
+  takeOffsets(offsets: number[]): TakeQuery {
+    return new TakeQuery(this.inner.takeOffsets(offsets));
+  }
+
+  takeRowIds(rowIds: number[]): TakeQuery {
+    return new TakeQuery(this.inner.takeRowIds(rowIds));
+  }
+
   query(): Query {
     return new Query(this.inner);
   }

nodejs/npm/darwin-arm64/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-darwin-arm64",
-	"version": "0.21.2",
+	"version": "0.22.1-beta.3",
 	"os": ["darwin"],
 	"cpu": ["arm64"],
 	"main": "lancedb.darwin-arm64.node",

nodejs/npm/darwin-x64/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-darwin-x64",
-	"version": "0.21.2",
+	"version": "0.22.1-beta.3",
 	"os": ["darwin"],
 	"cpu": ["x64"],
 	"main": "lancedb.darwin-x64.node",

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

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-arm64-gnu",
-	"version": "0.21.2",
+	"version": "0.22.1-beta.3",
 	"os": ["linux"],
 	"cpu": ["arm64"],
 	"main": "lancedb.linux-arm64-gnu.node",

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

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-arm64-musl",
-	"version": "0.21.2",
+	"version": "0.22.1-beta.3",
 	"os": ["linux"],
 	"cpu": ["arm64"],
 	"main": "lancedb.linux-arm64-musl.node",

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

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-x64-gnu",
-	"version": "0.21.2",
+	"version": "0.22.1-beta.3",
 	"os": ["linux"],
 	"cpu": ["x64"],
 	"main": "lancedb.linux-x64-gnu.node",

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

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-x64-musl",
-	"version": "0.21.2",
+	"version": "0.22.1-beta.3",
 	"os": ["linux"],
 	"cpu": ["x64"],
 	"main": "lancedb.linux-x64-musl.node",

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

@@ -1,6 +1,6 @@
 {
   "name": "@lancedb/lancedb-win32-arm64-msvc",
-  "version": "0.21.2",
+  "version": "0.22.1-beta.3",
   "os": [
     "win32"
   ],

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

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-win32-x64-msvc",
-	"version": "0.21.2",
+	"version": "0.22.1-beta.3",
 	"os": ["win32"],
 	"cpu": ["x64"],
 	"main": "lancedb.win32-x64-msvc.node",

nodejs/package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "@lancedb/lancedb",
-  "version": "0.21.2",
+  "version": "0.22.1-beta.3",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "@lancedb/lancedb",
-      "version": "0.21.2",
+      "version": "0.22.1-beta.3",
       "cpu": [
         "x64",
         "arm64"
@@ -5549,10 +5549,11 @@
       "dev": true
     },
     "node_modules/brace-expansion": {
-      "version": "1.1.11",
-      "resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-1.1.11.tgz",
-      "integrity": "sha512-iCuPHDFgrHX7H2vEI/5xpz07zSHB00TpugqhmYtVmMO6518mCuRMoOYFldEBl0g187ufozdaHgWKcYFb61qGiA==",
+      "version": "1.1.12",
+      "resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-1.1.12.tgz",
+      "integrity": "sha512-9T9UjW3r0UW5c1Q7GTwllptXwhvYmEzFhzMfZ9H7FQWt+uZePjZPjBP/W1ZEyZ1twGWom5/56TF4lPcqjnDHcg==",
       "dev": true,
+      "license": "MIT",
       "dependencies": {
         "balanced-match": "^1.0.0",
         "concat-map": "0.0.1"
@@ -5629,6 +5630,20 @@
       "integrity": "sha512-E+XQCRwSbaaiChtv6k6Dwgc+bx+Bs6vuKJHHl5kox/BaKbhiXzqQOwK4cO22yElGp2OCmjwVhT3HmxgyPGnJfQ==",
       "dev": true
     },
+    "node_modules/call-bind-apply-helpers": {
+      "version": "1.0.2",
+      "resolved": "https://registry.npmjs.org/call-bind-apply-helpers/-/call-bind-apply-helpers-1.0.2.tgz",
+      "integrity": "sha512-Sp1ablJ0ivDkSzjcaJdxEunN5/XvksFJ2sMBFfq6x0ryhQV/2b/KwFe21cMpmHtPOSij8K99/wSfoEuTObmuMQ==",
+      "devOptional": true,
+      "license": "MIT",
+      "dependencies": {
+        "es-errors": "^1.3.0",
+        "function-bind": "^1.1.2"
+      },
+      "engines": {
+        "node": ">= 0.4"
+      }
+    },
     "node_modules/camelcase": {
       "version": "5.3.1",
       "resolved": "https://registry.npmjs.org/camelcase/-/camelcase-5.3.1.tgz",
@@ -6032,6 +6047,21 @@
         "node": ">=6.0.0"
       }
     },
+    "node_modules/dunder-proto": {
+      "version": "1.0.1",
+      "resolved": "https://registry.npmjs.org/dunder-proto/-/dunder-proto-1.0.1.tgz",
+      "integrity": "sha512-KIN/nDJBQRcXw0MLVhZE9iQHmG68qAVIBg9CqmUYjmQIhgij9U5MFvrqkUL5FbtyyzZuOeOt0zdeRe4UY7ct+A==",
+      "devOptional": true,
+      "license": "MIT",
+      "dependencies": {
+        "call-bind-apply-helpers": "^1.0.1",
+        "es-errors": "^1.3.0",
+        "gopd": "^1.2.0"
+      },
+      "engines": {
+        "node": ">= 0.4"
+      }
+    },
     "node_modules/eastasianwidth": {
       "version": "0.2.0",
       "resolved": "https://registry.npmjs.org/eastasianwidth/-/eastasianwidth-0.2.0.tgz",
@@ -6071,6 +6101,55 @@
         "is-arrayish": "^0.2.1"
       }
     },
+    "node_modules/es-define-property": {
+      "version": "1.0.1",
+      "resolved": "https://registry.npmjs.org/es-define-property/-/es-define-property-1.0.1.tgz",
+      "integrity": "sha512-e3nRfgfUZ4rNGL232gUgX06QNyyez04KdjFrF+LTRoOXmrOgFKDg4BCdsjW8EnT69eqdYGmRpJwiPVYNrCaW3g==",
+      "devOptional": true,
+      "license": "MIT",
+      "engines": {
+        "node": ">= 0.4"
+      }
+    },
+    "node_modules/es-errors": {
+      "version": "1.3.0",
+      "resolved": "https://registry.npmjs.org/es-errors/-/es-errors-1.3.0.tgz",
+      "integrity": "sha512-Zf5H2Kxt2xjTvbJvP2ZWLEICxA6j+hAmMzIlypy4xcBg1vKVnx89Wy0GbS+kf5cwCVFFzdCFh2XSCFNULS6csw==",
+      "devOptional": true,
+      "license": "MIT",
+      "engines": {
+        "node": ">= 0.4"
+      }
+    },
+    "node_modules/es-object-atoms": {
+      "version": "1.1.1",
+      "resolved": "https://registry.npmjs.org/es-object-atoms/-/es-object-atoms-1.1.1.tgz",
+      "integrity": "sha512-FGgH2h8zKNim9ljj7dankFPcICIK9Cp5bm+c2gQSYePhpaG5+esrLODihIorn+Pe6FGJzWhXQotPv73jTaldXA==",
+      "devOptional": true,
+      "license": "MIT",
+      "dependencies": {
+        "es-errors": "^1.3.0"
+      },
+      "engines": {
+        "node": ">= 0.4"
+      }
+    },
+    "node_modules/es-set-tostringtag": {
+      "version": "2.1.0",
+      "resolved": "https://registry.npmjs.org/es-set-tostringtag/-/es-set-tostringtag-2.1.0.tgz",
+      "integrity": "sha512-j6vWzfrGVfyXxge+O0x5sh6cvxAog0a/4Rdd2K36zCMV5eJ+/+tOAngRO8cODMNWbVRdVlmGZQL2YS3yR8bIUA==",
+      "devOptional": true,
+      "license": "MIT",
+      "dependencies": {
+        "es-errors": "^1.3.0",
+        "get-intrinsic": "^1.2.6",
+        "has-tostringtag": "^1.0.2",
+        "hasown": "^2.0.2"
+      },
+      "engines": {
+        "node": ">= 0.4"
+      }
+    },
     "node_modules/escalade": {
       "version": "3.1.1",
       "resolved": "https://registry.npmjs.org/escalade/-/escalade-3.1.1.tgz",
@@ -6510,13 +6589,16 @@
       }
     },
     "node_modules/form-data": {
-      "version": "4.0.0",
-      "resolved": "https://registry.npmjs.org/form-data/-/form-data-4.0.0.tgz",
-      "integrity": "sha512-ETEklSGi5t0QMZuiXoA/Q6vcnxcLQP5vdugSpuAyi6SVGi2clPPp+xgEhuMaHC+zGgn31Kd235W35f7Hykkaww==",
+      "version": "4.0.4",
+      "resolved": "https://registry.npmjs.org/form-data/-/form-data-4.0.4.tgz",
+      "integrity": "sha512-KrGhL9Q4zjj0kiUt5OO4Mr/A/jlI2jDYs5eHBpYHPcBEVSiipAvn2Ko2HnPe20rmcuuvMHNdZFp+4IlGTMF0Ow==",
       "devOptional": true,
+      "license": "MIT",
       "dependencies": {
         "asynckit": "^0.4.0",
         "combined-stream": "^1.0.8",
+        "es-set-tostringtag": "^2.1.0",
+        "hasown": "^2.0.2",
         "mime-types": "^2.1.12"
       },
       "engines": {
@@ -6575,7 +6657,7 @@
       "version": "1.1.2",
       "resolved": "https://registry.npmjs.org/function-bind/-/function-bind-1.1.2.tgz",
       "integrity": "sha512-7XHNxH7qX9xG5mIwxkhumTox/MIRNcOgDrxWsMt2pAr23WHp6MrRlN7FBSFpCpr+oVO0F744iUgR82nJMfG2SA==",
-      "dev": true,
+      "devOptional": true,
       "funding": {
         "url": "https://github.com/sponsors/ljharb"
       }
@@ -6598,6 +6680,31 @@
         "node": "6.* || 8.* || >= 10.*"
       }
     },
+    "node_modules/get-intrinsic": {
+      "version": "1.3.0",
+      "resolved": "https://registry.npmjs.org/get-intrinsic/-/get-intrinsic-1.3.0.tgz",
+      "integrity": "sha512-9fSjSaos/fRIVIp+xSJlE6lfwhES7LNtKaCBIamHsjr2na1BiABJPo0mOjjz8GJDURarmCPGqaiVg5mfjb98CQ==",
+      "devOptional": true,
+      "license": "MIT",
+      "dependencies": {
+        "call-bind-apply-helpers": "^1.0.2",
+        "es-define-property": "^1.0.1",
+        "es-errors": "^1.3.0",
+        "es-object-atoms": "^1.1.1",
+        "function-bind": "^1.1.2",
+        "get-proto": "^1.0.1",
+        "gopd": "^1.2.0",
+        "has-symbols": "^1.1.0",
+        "hasown": "^2.0.2",
+        "math-intrinsics": "^1.1.0"
+      },
+      "engines": {
+        "node": ">= 0.4"
+      },
+      "funding": {
+        "url": "https://github.com/sponsors/ljharb"
+      }
+    },
     "node_modules/get-package-type": {
       "version": "0.1.0",
       "resolved": "https://registry.npmjs.org/get-package-type/-/get-package-type-0.1.0.tgz",
@@ -6607,6 +6714,20 @@
         "node": ">=8.0.0"
       }
     },
+    "node_modules/get-proto": {
+      "version": "1.0.1",
+      "resolved": "https://registry.npmjs.org/get-proto/-/get-proto-1.0.1.tgz",
+      "integrity": "sha512-sTSfBjoXBp89JvIKIefqw7U2CCebsc74kiY6awiGogKtoSGbgjYE/G/+l9sF3MWFPNc9IcoOC4ODfKHfxFmp0g==",
+      "devOptional": true,
+      "license": "MIT",
+      "dependencies": {
+        "dunder-proto": "^1.0.1",
+        "es-object-atoms": "^1.0.0"
+      },
+      "engines": {
+        "node": ">= 0.4"
+      }
+    },
     "node_modules/get-stream": {
       "version": "6.0.1",
       "resolved": "https://registry.npmjs.org/get-stream/-/get-stream-6.0.1.tgz",
@@ -6698,6 +6819,19 @@
         "url": "https://github.com/sponsors/sindresorhus"
       }
     },
+    "node_modules/gopd": {
+      "version": "1.2.0",
+      "resolved": "https://registry.npmjs.org/gopd/-/gopd-1.2.0.tgz",
+      "integrity": "sha512-ZUKRh6/kUFoAiTAtTYPZJ3hw9wNxx+BIBOijnlG9PnrJsCcSjs1wyyD6vJpaYtgnzDrKYRSqf3OO6Rfa93xsRg==",
+      "devOptional": true,
+      "license": "MIT",
+      "engines": {
+        "node": ">= 0.4"
+      },
+      "funding": {
+        "url": "https://github.com/sponsors/ljharb"
+      }
+    },
     "node_modules/graceful-fs": {
       "version": "4.2.11",
       "resolved": "https://registry.npmjs.org/graceful-fs/-/graceful-fs-4.2.11.tgz",
@@ -6724,11 +6858,41 @@
         "node": ">=8"
       }
     },
+    "node_modules/has-symbols": {
+      "version": "1.1.0",
+      "resolved": "https://registry.npmjs.org/has-symbols/-/has-symbols-1.1.0.tgz",
+      "integrity": "sha512-1cDNdwJ2Jaohmb3sg4OmKaMBwuC48sYni5HUw2DvsC8LjGTLK9h+eb1X6RyuOHe4hT0ULCW68iomhjUoKUqlPQ==",
+      "devOptional": true,
+      "license": "MIT",
+      "engines": {
+        "node": ">= 0.4"
+      },
+      "funding": {
+        "url": "https://github.com/sponsors/ljharb"
+      }
+    },
+    "node_modules/has-tostringtag": {
+      "version": "1.0.2",
+      "resolved": "https://registry.npmjs.org/has-tostringtag/-/has-tostringtag-1.0.2.tgz",
+      "integrity": "sha512-NqADB8VjPFLM2V0VvHUewwwsw0ZWBaIdgo+ieHtK3hasLz4qeCRjYcqfB6AQrBggRKppKF8L52/VqdVsO47Dlw==",
+      "devOptional": true,
+      "license": "MIT",
+      "dependencies": {
+        "has-symbols": "^1.0.3"
+      },
+      "engines": {
+        "node": ">= 0.4"
+      },
+      "funding": {
+        "url": "https://github.com/sponsors/ljharb"
+      }
+    },
     "node_modules/hasown": {
-      "version": "2.0.0",
-      "resolved": "https://registry.npmjs.org/hasown/-/hasown-2.0.0.tgz",
-      "integrity": "sha512-vUptKVTpIJhcczKBbgnS+RtcuYMB8+oNzPK2/Hp3hanz8JmpATdmmgLgSaadVREkDm+e2giHwY3ZRkyjSIDDFA==",
-      "dev": true,
+      "version": "2.0.2",
+      "resolved": "https://registry.npmjs.org/hasown/-/hasown-2.0.2.tgz",
+      "integrity": "sha512-0hJU9SCPvmMzIBdZFqNPXWa6dqh7WdH0cII9y+CyS8rG3nL48Bclra9HmKhVVUHyPWNH5Y7xDwAB7bfgSjkUMQ==",
+      "devOptional": true,
+      "license": "MIT",
       "dependencies": {
         "function-bind": "^1.1.2"
       },
@@ -7943,6 +8107,16 @@
       "integrity": "sha512-8+9WqebbFzpX9OR+Wa6O29asIogeRMzcGtAINdpMHHyAg10f05aSFVBbcEqGf/PXw1EjAZ+q2/bEBg3DvurK3Q==",
       "dev": true
     },
+    "node_modules/math-intrinsics": {
+      "version": "1.1.0",
+      "resolved": "https://registry.npmjs.org/math-intrinsics/-/math-intrinsics-1.1.0.tgz",
+      "integrity": "sha512-/IXtbwEk5HTPyEwyKX6hGkYXxM9nbj64B+ilVJnC/R6B0pH5G4V3b0pVbL7DBj4tkhBAppbQUlf6F6Xl9LHu1g==",
+      "devOptional": true,
+      "license": "MIT",
+      "engines": {
+        "node": ">= 0.4"
+      }
+    },
     "node_modules/md5": {
       "version": "2.3.0",
       "resolved": "https://registry.npmjs.org/md5/-/md5-2.3.0.tgz",
@@ -8053,9 +8227,10 @@
       }
     },
     "node_modules/minizlib/node_modules/brace-expansion": {
-      "version": "2.0.1",
-      "resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-2.0.1.tgz",
-      "integrity": "sha512-XnAIvQ8eM+kC6aULx6wuQiwVsnzsi9d3WxzV3FpWTGA19F621kwdbsAcFKXgKUHZWsy+mY6iL1sHTxWEFCytDA==",
+      "version": "2.0.2",
+      "resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-2.0.2.tgz",
+      "integrity": "sha512-Jt0vHyM+jmUBqojB7E1NIYadt0vI0Qxjxd2TErW94wDz+E2LAm5vKMXXwg6ZZBTHPuUlDgQHKXvjGBdfcF1ZDQ==",
+      "license": "MIT",
       "optional": true,
       "dependencies": {
         "balanced-match": "^1.0.0"
@@ -9201,10 +9376,11 @@
       "dev": true
     },
     "node_modules/tmp": {
-      "version": "0.2.3",
-      "resolved": "https://registry.npmjs.org/tmp/-/tmp-0.2.3.tgz",
-      "integrity": "sha512-nZD7m9iCPC5g0pYmcaxogYKggSfLsdxl8of3Q/oIbqCqLLIO9IAF0GWjX1z9NZRHPiXv8Wex4yDCaZsgEw0Y8w==",
+      "version": "0.2.5",
+      "resolved": "https://registry.npmjs.org/tmp/-/tmp-0.2.5.tgz",
+      "integrity": "sha512-voyz6MApa1rQGUxT3E+BK7/ROe8itEx7vD8/HEvt4xwXucvQ5G5oeEiHkmHZJuBO21RpOf+YYm9MOivj709jow==",
       "dev": true,
+      "license": "MIT",
       "engines": {
         "node": ">=14.14"
       }
@@ -9349,10 +9525,11 @@
       }
     },
     "node_modules/typedoc/node_modules/brace-expansion": {
-      "version": "2.0.1",
-      "resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-2.0.1.tgz",
-      "integrity": "sha512-XnAIvQ8eM+kC6aULx6wuQiwVsnzsi9d3WxzV3FpWTGA19F621kwdbsAcFKXgKUHZWsy+mY6iL1sHTxWEFCytDA==",
+      "version": "2.0.2",
+      "resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-2.0.2.tgz",
+      "integrity": "sha512-Jt0vHyM+jmUBqojB7E1NIYadt0vI0Qxjxd2TErW94wDz+E2LAm5vKMXXwg6ZZBTHPuUlDgQHKXvjGBdfcF1ZDQ==",
       "dev": true,
+      "license": "MIT",
       "dependencies": {
         "balanced-match": "^1.0.0"
       }
@@ -9602,10 +9779,11 @@
       }
     },
     "node_modules/typescript-eslint/node_modules/brace-expansion": {
-      "version": "2.0.1",
-      "resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-2.0.1.tgz",
-      "integrity": "sha512-XnAIvQ8eM+kC6aULx6wuQiwVsnzsi9d3WxzV3FpWTGA19F621kwdbsAcFKXgKUHZWsy+mY6iL1sHTxWEFCytDA==",
+      "version": "2.0.2",
+      "resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-2.0.2.tgz",
+      "integrity": "sha512-Jt0vHyM+jmUBqojB7E1NIYadt0vI0Qxjxd2TErW94wDz+E2LAm5vKMXXwg6ZZBTHPuUlDgQHKXvjGBdfcF1ZDQ==",
       "dev": true,
+      "license": "MIT",
       "dependencies": {
         "balanced-match": "^1.0.0"
       }

nodejs/package.json

@@ -11,7 +11,7 @@
     "ann"
   ],
   "private": false,
-  "version": "0.21.2",
+  "version": "0.22.1-beta.3",
   "main": "dist/index.js",
   "exports": {
     ".": "./dist/index.js",

nodejs/src/connection.rs

@@ -2,12 +2,14 @@
 // SPDX-FileCopyrightText: Copyright The LanceDB Authors
 
 use std::collections::HashMap;
+use std::sync::Arc;
 
 use lancedb::database::CreateTableMode;
 use napi::bindgen_prelude::*;
 use napi_derive::*;
 
 use crate::error::NapiErrorExt;
+use crate::header::JsHeaderProvider;
 use crate::table::Table;
 use crate::ConnectionOptions;
 use lancedb::connection::{ConnectBuilder, Connection as LanceDBConnection};
@@ -45,7 +47,11 @@ impl Connection {
 impl Connection {
     /// Create a new Connection instance from the given URI.
     #[napi(factory)]
-    pub async fn new(uri: String, options: ConnectionOptions) -> napi::Result<Self> {
+    pub async fn new(
+        uri: String,
+        options: ConnectionOptions,
+        header_provider: Option<&JsHeaderProvider>,
+    ) -> napi::Result<Self> {
         let mut builder = ConnectBuilder::new(&uri);
         if let Some(interval) = options.read_consistency_interval {
             builder =
@@ -57,8 +63,16 @@ impl Connection {
             }
         }
 
+        // Create client config, optionally with header provider
         let client_config = options.client_config.unwrap_or_default();
-        builder = builder.client_config(client_config.into());
+        let mut rust_config: lancedb::remote::ClientConfig = client_config.into();
+
+        if let Some(provider) = header_provider {
+            rust_config.header_provider =
+                Some(Arc::new(provider.clone()) as Arc<dyn lancedb::remote::HeaderProvider>);
+        }
+
+        builder = builder.client_config(rust_config);
 
         if let Some(api_key) = options.api_key {
             builder = builder.api_key(&api_key);
@@ -100,10 +114,12 @@ impl Connection {
     #[napi(catch_unwind)]
     pub async fn table_names(
         &self,
+        namespace: Vec<String>,
         start_after: Option<String>,
         limit: Option<u32>,
     ) -> napi::Result<Vec<String>> {
         let mut op = self.get_inner()?.table_names();
+        op = op.namespace(namespace);
         if let Some(start_after) = start_after {
             op = op.start_after(start_after);
         }
@@ -125,6 +141,7 @@ impl Connection {
         name: String,
         buf: Buffer,
         mode: String,
+        namespace: Vec<String>,
         storage_options: Option<HashMap<String, String>>,
     ) -> napi::Result<Table> {
         let batches = ipc_file_to_batches(buf.to_vec())
@@ -132,6 +149,8 @@ impl Connection {
         let mode = Self::parse_create_mode_str(&mode)?;
         let mut builder = self.get_inner()?.create_table(&name, batches).mode(mode);
 
+        builder = builder.namespace(namespace);
+
         if let Some(storage_options) = storage_options {
             for (key, value) in storage_options {
                 builder = builder.storage_option(key, value);
@@ -147,6 +166,7 @@ impl Connection {
         name: String,
         schema_buf: Buffer,
         mode: String,
+        namespace: Vec<String>,
         storage_options: Option<HashMap<String, String>>,
     ) -> napi::Result<Table> {
         let schema = ipc_file_to_schema(schema_buf.to_vec()).map_err(|e| {
@@ -157,6 +177,9 @@ impl Connection {
             .get_inner()?
             .create_empty_table(&name, schema)
             .mode(mode);
+
+        builder = builder.namespace(namespace);
+
         if let Some(storage_options) = storage_options {
             for (key, value) in storage_options {
                 builder = builder.storage_option(key, value);
@@ -170,10 +193,14 @@ impl Connection {
     pub async fn open_table(
         &self,
         name: String,
+        namespace: Vec<String>,
         storage_options: Option<HashMap<String, String>>,
         index_cache_size: Option<u32>,
     ) -> napi::Result<Table> {
         let mut builder = self.get_inner()?.open_table(&name);
+
+        builder = builder.namespace(namespace);
+
         if let Some(storage_options) = storage_options {
             for (key, value) in storage_options {
                 builder = builder.storage_option(key, value);
@@ -186,14 +213,50 @@ impl Connection {
         Ok(Table::new(tbl))
     }
 
+    #[napi(catch_unwind)]
+    pub async fn clone_table(
+        &self,
+        target_table_name: String,
+        source_uri: String,
+        target_namespace: Vec<String>,
+        source_version: Option<i64>,
+        source_tag: Option<String>,
+        is_shallow: bool,
+    ) -> napi::Result<Table> {
+        let mut builder = self
+            .get_inner()?
+            .clone_table(&target_table_name, &source_uri);
+
+        builder = builder.target_namespace(target_namespace);
+
+        if let Some(version) = source_version {
+            builder = builder.source_version(version as u64);
+        }
+
+        if let Some(tag) = source_tag {
+            builder = builder.source_tag(tag);
+        }
+
+        builder = builder.is_shallow(is_shallow);
+
+        let tbl = builder.execute().await.default_error()?;
+        Ok(Table::new(tbl))
+    }
+
     /// Drop table with the name. Or raise an error if the table does not exist.
     #[napi(catch_unwind)]
-    pub async fn drop_table(&self, name: String) -> napi::Result<()> {
-        self.get_inner()?.drop_table(&name).await.default_error()
+    pub async fn drop_table(&self, name: String, namespace: Vec<String>) -> napi::Result<()> {
+        self.get_inner()?
+            .drop_table(&name, &namespace)
+            .await
+            .default_error()
     }
 
     #[napi(catch_unwind)]
-    pub async fn drop_all_tables(&self) -> napi::Result<()> {
-        self.get_inner()?.drop_all_tables().await.default_error()
+    pub async fn drop_all_tables(&self, namespace: Vec<String>) -> napi::Result<()> {
+        self.get_inner()?
+            .drop_all_tables(&namespace)
+            .await
+            .default_error()
     }
 }

nodejs/src/header.rs

@@ -0,0 +1,71 @@
+// SPDX-License-Identifier: Apache-2.0
+// SPDX-FileCopyrightText: Copyright The LanceDB Authors
+
+use napi::{
+    bindgen_prelude::*,
+    threadsafe_function::{ErrorStrategy, ThreadsafeFunction},
+};
+use napi_derive::napi;
+use std::collections::HashMap;
+use std::sync::Arc;
+
+/// JavaScript HeaderProvider implementation that wraps a JavaScript callback.
+/// This is the only native header provider - all header provider implementations
+/// should provide a JavaScript function that returns headers.
+#[napi]
+pub struct JsHeaderProvider {
+    get_headers_fn: Arc<ThreadsafeFunction<(), ErrorStrategy::CalleeHandled>>,
+}
+
+impl Clone for JsHeaderProvider {
+    fn clone(&self) -> Self {
+        Self {
+            get_headers_fn: self.get_headers_fn.clone(),
+        }
+    }
+}
+
+#[napi]
+impl JsHeaderProvider {
+    /// Create a new JsHeaderProvider from a JavaScript callback
+    #[napi(constructor)]
+    pub fn new(get_headers_callback: JsFunction) -> Result<Self> {
+        let get_headers_fn = get_headers_callback
+            .create_threadsafe_function(0, |ctx| Ok(vec![ctx.value]))
+            .map_err(|e| {
+                Error::new(
+                    Status::GenericFailure,
+                    format!("Failed to create threadsafe function: {}", e),
+                )
+            })?;
+
+        Ok(Self {
+            get_headers_fn: Arc::new(get_headers_fn),
+        })
+    }
+}
+
+#[cfg(feature = "remote")]
+#[async_trait::async_trait]
+impl lancedb::remote::HeaderProvider for JsHeaderProvider {
+    async fn get_headers(&self) -> lancedb::error::Result<HashMap<String, String>> {
+        // Call the JavaScript function asynchronously
+        let promise: Promise<HashMap<String, String>> =
+            self.get_headers_fn.call_async(Ok(())).await.map_err(|e| {
+                lancedb::error::Error::Runtime {
+                    message: format!("Failed to call JavaScript get_headers: {}", e),
+                }
+            })?;
+
+        // Await the promise result
+        promise.await.map_err(|e| lancedb::error::Error::Runtime {
+            message: format!("JavaScript get_headers failed: {}", e),
+        })
+    }
+}
+
+impl std::fmt::Debug for JsHeaderProvider {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        write!(f, "JsHeaderProvider")
+    }
+}

nodejs/src/lib.rs

@@ -8,6 +8,7 @@ use napi_derive::*;
 
 mod connection;
 mod error;
+mod header;
 mod index;
 mod iterator;
 pub mod merge;

nodejs/src/query.rs

@@ -12,6 +12,7 @@ use lancedb::query::Query as LanceDbQuery;
 use lancedb::query::QueryBase;
 use lancedb::query::QueryExecutionOptions;
 use lancedb::query::Select;
+use lancedb::query::TakeQuery as LanceDbTakeQuery;
 use lancedb::query::VectorQuery as LanceDbVectorQuery;
 use napi::bindgen_prelude::*;
 use napi_derive::napi;
@@ -319,6 +320,79 @@ impl VectorQuery {
     }
 }
 
+#[napi]
+pub struct TakeQuery {
+    inner: LanceDbTakeQuery,
+}
+
+#[napi]
+impl TakeQuery {
+    pub fn new(query: LanceDbTakeQuery) -> Self {
+        Self { inner: query }
+    }
+
+    #[napi]
+    pub fn select(&mut self, columns: Vec<(String, String)>) {
+        self.inner = self.inner.clone().select(Select::dynamic(&columns));
+    }
+
+    #[napi]
+    pub fn select_columns(&mut self, columns: Vec<String>) {
+        self.inner = self.inner.clone().select(Select::columns(&columns));
+    }
+
+    #[napi]
+    pub fn with_row_id(&mut self) {
+        self.inner = self.inner.clone().with_row_id();
+    }
+
+    #[napi(catch_unwind)]
+    pub async fn execute(
+        &self,
+        max_batch_length: Option<u32>,
+        timeout_ms: Option<u32>,
+    ) -> napi::Result<RecordBatchIterator> {
+        let mut execution_opts = QueryExecutionOptions::default();
+        if let Some(max_batch_length) = max_batch_length {
+            execution_opts.max_batch_length = max_batch_length;
+        }
+        if let Some(timeout_ms) = timeout_ms {
+            execution_opts.timeout = Some(std::time::Duration::from_millis(timeout_ms as u64))
+        }
+        let inner_stream = self
+            .inner
+            .execute_with_options(execution_opts)
+            .await
+            .map_err(|e| {
+                napi::Error::from_reason(format!(
+                    "Failed to execute query stream: {}",
+                    convert_error(&e)
+                ))
+            })?;
+        Ok(RecordBatchIterator::new(inner_stream))
+    }
+
+    #[napi]
+    pub async fn explain_plan(&self, verbose: bool) -> napi::Result<String> {
+        self.inner.explain_plan(verbose).await.map_err(|e| {
+            napi::Error::from_reason(format!(
+                "Failed to retrieve the query plan: {}",
+                convert_error(&e)
+            ))
+        })
+    }
+
+    #[napi(catch_unwind)]
+    pub async fn analyze_plan(&self) -> napi::Result<String> {
+        self.inner.analyze_plan().await.map_err(|e| {
+            napi::Error::from_reason(format!(
+                "Failed to execute analyze plan: {}",
+                convert_error(&e)
+            ))
+        })
+    }
+}
+
 #[napi]
 #[derive(Debug, Clone)]
 pub struct JsFullTextQuery {
@@ -406,6 +480,7 @@ impl JsFullTextQuery {
     }
 
     #[napi(factory)]
+    #[allow(clippy::use_self)] // NAPI doesn't allow Self here but clippy reports it
     pub fn boolean_query(queries: Vec<(String, &JsFullTextQuery)>) -> napi::Result<Self> {
         let mut sub_queries = Vec::with_capacity(queries.len());
         for (occur, q) in queries {

nodejs/src/remote.rs

@@ -69,6 +69,20 @@ pub struct RetryConfig {
     pub statuses: Option<Vec<u16>>,
 }
 
+/// TLS/mTLS configuration for the remote HTTP client.
+#[napi(object)]
+#[derive(Debug, Default)]
+pub struct TlsConfig {
+    /// Path to the client certificate file (PEM format) for mTLS authentication.
+    pub cert_file: Option<String>,
+    /// Path to the client private key file (PEM format) for mTLS authentication.
+    pub key_file: Option<String>,
+    /// Path to the CA certificate file (PEM format) for server verification.
+    pub ssl_ca_cert: Option<String>,
+    /// Whether to verify the hostname in the server's certificate.
+    pub assert_hostname: Option<bool>,
+}
+
 #[napi(object)]
 #[derive(Debug, Default)]
 pub struct ClientConfig {
@@ -76,6 +90,8 @@ pub struct ClientConfig {
     pub retry_config: Option<RetryConfig>,
     pub timeout_config: Option<TimeoutConfig>,
     pub extra_headers: Option<HashMap<String, String>>,
+    pub id_delimiter: Option<String>,
+    pub tls_config: Option<TlsConfig>,
 }
 
 impl From<TimeoutConfig> for lancedb::remote::TimeoutConfig {
@@ -106,6 +122,17 @@ impl From<RetryConfig> for lancedb::remote::RetryConfig {
     }
 }
 
+impl From<TlsConfig> for lancedb::remote::TlsConfig {
+    fn from(config: TlsConfig) -> Self {
+        Self {
+            cert_file: config.cert_file,
+            key_file: config.key_file,
+            ssl_ca_cert: config.ssl_ca_cert,
+            assert_hostname: config.assert_hostname.unwrap_or(true),
+        }
+    }
+}
+
 impl From<ClientConfig> for lancedb::remote::ClientConfig {
     fn from(config: ClientConfig) -> Self {
         Self {
@@ -115,6 +142,9 @@ impl From<ClientConfig> for lancedb::remote::ClientConfig {
             retry_config: config.retry_config.map(Into::into).unwrap_or_default(),
             timeout_config: config.timeout_config.map(Into::into).unwrap_or_default(),
             extra_headers: config.extra_headers.unwrap_or_default(),
+            id_delimiter: config.id_delimiter,
+            tls_config: config.tls_config.map(Into::into),
+            header_provider: None, // the header provider is set separately later
         }
     }
 }

nodejs/src/session.rs

@@ -94,7 +94,7 @@ impl napi::bindgen_prelude::FromNapiValue for Session {
         env: napi::sys::napi_env,
         napi_val: napi::sys::napi_value,
     ) -> napi::Result<Self> {
-        let object: napi::bindgen_prelude::ClassInstance<Session> =
+        let object: napi::bindgen_prelude::ClassInstance<Self> =
             napi::bindgen_prelude::ClassInstance::from_napi_value(env, napi_val)?;
         let copy = object.clone();
         Ok(copy)

nodejs/src/table.rs

@@ -15,7 +15,7 @@ use napi_derive::napi;
 use crate::error::NapiErrorExt;
 use crate::index::Index;
 use crate::merge::NativeMergeInsertBuilder;
-use crate::query::{Query, VectorQuery};
+use crate::query::{Query, TakeQuery, VectorQuery};
 
 #[napi]
 pub struct Table {
@@ -114,6 +114,8 @@ impl Table {
         column: String,
         replace: Option<bool>,
         wait_timeout_s: Option<i64>,
+        name: Option<String>,
+        train: Option<bool>,
     ) -> napi::Result<()> {
         let lancedb_index = if let Some(index) = index {
             index.consume()?
@@ -128,6 +130,12 @@ impl Table {
             builder =
                 builder.wait_timeout(std::time::Duration::from_secs(timeout.try_into().unwrap()));
         }
+        if let Some(name) = name {
+            builder = builder.name(name);
+        }
+        if let Some(train) = train {
+            builder = builder.train(train);
+        }
         builder.execute().await.default_error()
     }
 
@@ -187,6 +195,44 @@ impl Table {
         Ok(Query::new(self.inner_ref()?.query()))
     }
 
+    #[napi(catch_unwind)]
+    pub fn take_offsets(&self, offsets: Vec<i64>) -> napi::Result<TakeQuery> {
+        Ok(TakeQuery::new(
+            self.inner_ref()?.take_offsets(
+                offsets
+                    .into_iter()
+                    .map(|o| {
+                        u64::try_from(o).map_err(|e| {
+                            napi::Error::from_reason(format!(
+                                "Failed to convert offset to u64: {}",
+                                e
+                            ))
+                        })
+                    })
+                    .collect::<Result<Vec<_>>>()?,
+            ),
+        ))
+    }
+
+    #[napi(catch_unwind)]
+    pub fn take_row_ids(&self, row_ids: Vec<i64>) -> napi::Result<TakeQuery> {
+        Ok(TakeQuery::new(
+            self.inner_ref()?.take_row_ids(
+                row_ids
+                    .into_iter()
+                    .map(|o| {
+                        u64::try_from(o).map_err(|e| {
+                            napi::Error::from_reason(format!(
+                                "Failed to convert row id to u64: {}",
+                                e
+                            ))
+                        })
+                    })
+                    .collect::<Result<Vec<_>>>()?,
+            ),
+        ))
+    }
+
     #[napi(catch_unwind)]
     pub fn vector_search(&self, vector: Float32Array) -> napi::Result<VectorQuery> {
         self.query()?.nearest_to(vector)

nodejs/typedoc.json

@@ -2,6 +2,7 @@
   "intentionallyNotExported": [
     "lancedb/native.d.ts:Query",
     "lancedb/native.d.ts:VectorQuery",
+    "lancedb/native.d.ts:TakeQuery",
     "lancedb/native.d.ts:RecordBatchIterator",
     "lancedb/native.d.ts:NativeMergeInsertBuilder"
   ],

python/.bumpversion.toml

@@ -1,5 +1,5 @@
 [tool.bumpversion]
-current_version = "0.24.2"
+current_version = "0.25.1"
 parse = """(?x)
     (?P<major>0|[1-9]\\d*)\\.
     (?P<minor>0|[1-9]\\d*)\\.

python/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "lancedb-python"
-version = "0.24.2"
+version = "0.25.1"
 edition.workspace = true
 description = "Python bindings for LanceDB"
 license.workspace = true
@@ -15,6 +15,7 @@ crate-type = ["cdylib"]
 
 [dependencies]
 arrow = { version = "55.1", features = ["pyarrow"] }
+async-trait = "0.1"
 lancedb = { path = "../rust/lancedb", default-features = false }
 env_logger.workspace = true
 pyo3 = { version = "0.24", features = ["extension-module", "abi3-py39"] }
@@ -33,6 +34,6 @@ pyo3-build-config = { version = "0.24", features = [
 ] }
 
 [features]
-default = ["remote"]
+default = ["remote", "lancedb/default"]
 fp16kernels = ["lancedb/fp16kernels"]
 remote = ["lancedb/remote"]

python/pyproject.toml

@@ -10,6 +10,7 @@ dependencies = [
     "pyarrow>=16",
     "pydantic>=1.10",
     "tqdm>=4.27.0",
+    "lance-namespace==0.0.6"
 ]
 description = "lancedb"
 authors = [{ name = "LanceDB Devs", email = "dev@lancedb.com" }]

python/python/lancedb/__init__.py

@@ -19,6 +19,7 @@ from .remote.db import RemoteDBConnection
 from .schema import vector
 from .table import AsyncTable
 from ._lancedb import Session
+from .namespace import connect_namespace, LanceNamespaceDBConnection
 
 
 def connect(
@@ -221,6 +222,7 @@ async def connect_async(
 __all__ = [
     "connect",
     "connect_async",
+    "connect_namespace",
     "AsyncConnection",
     "AsyncTable",
     "URI",
@@ -228,6 +230,7 @@ __all__ = [
     "vector",
     "DBConnection",
     "LanceDBConnection",
+    "LanceNamespaceDBConnection",
     "RemoteDBConnection",
     "Session",
     "__version__",

python/python/lancedb/_lancedb.pyi

@@ -21,14 +21,28 @@ class Session:
 
 class Connection(object):
     uri: str
+    async def is_open(self): ...
+    async def close(self): ...
+    async def list_namespaces(
+        self,
+        namespace: List[str],
+        page_token: Optional[str],
+        limit: Optional[int],
+    ) -> List[str]: ...
+    async def create_namespace(self, namespace: List[str]) -> None: ...
+    async def drop_namespace(self, namespace: List[str]) -> None: ...
     async def table_names(
-        self, start_after: Optional[str], limit: Optional[int]
+        self,
+        namespace: List[str],
+        start_after: Optional[str],
+        limit: Optional[int],
     ) -> list[str]: ...
     async def create_table(
         self,
         name: str,
         mode: str,
         data: pa.RecordBatchReader,
+        namespace: List[str] = [],
         storage_options: Optional[Dict[str, str]] = None,
     ) -> Table: ...
     async def create_empty_table(
@@ -36,10 +50,34 @@ class Connection(object):
         name: str,
         mode: str,
         schema: pa.Schema,
+        namespace: List[str] = [],
         storage_options: Optional[Dict[str, str]] = None,
     ) -> Table: ...
-    async def rename_table(self, old_name: str, new_name: str) -> None: ...
-    async def drop_table(self, name: str) -> None: ...
+    async def open_table(
+        self,
+        name: str,
+        namespace: List[str] = [],
+        storage_options: Optional[Dict[str, str]] = None,
+        index_cache_size: Optional[int] = None,
+    ) -> Table: ...
+    async def clone_table(
+        self,
+        target_table_name: str,
+        source_uri: str,
+        target_namespace: List[str] = [],
+        source_version: Optional[int] = None,
+        source_tag: Optional[str] = None,
+        is_shallow: bool = True,
+    ) -> Table: ...
+    async def rename_table(
+        self,
+        cur_name: str,
+        new_name: str,
+        cur_namespace: List[str] = [],
+        new_namespace: List[str] = [],
+    ) -> None: ...
+    async def drop_table(self, name: str, namespace: List[str] = []) -> None: ...
+    async def drop_all_tables(self, namespace: List[str] = []) -> None: ...
 
 class Table:
     def name(self) -> str: ...
@@ -59,6 +97,10 @@ class Table:
         column: str,
         index: Union[IvfFlat, IvfPq, HnswPq, HnswSq, BTree, Bitmap, LabelList, FTS],
         replace: Optional[bool],
+        wait_timeout: Optional[object],
+        *,
+        name: Optional[str],
+        train: Optional[bool],
     ): ...
     async def list_versions(self) -> List[Dict[str, Any]]: ...
     async def version(self) -> int: ...

python/python/lancedb/db.py

@@ -43,14 +43,70 @@ if TYPE_CHECKING:
 class DBConnection(EnforceOverrides):
     """An active LanceDB connection interface."""
 
+    def list_namespaces(
+        self,
+        namespace: List[str] = [],
+        page_token: Optional[str] = None,
+        limit: int = 10,
+    ) -> Iterable[str]:
+        """List immediate child namespace names in the given namespace.
+
+        Parameters
+        ----------
+        namespace: List[str], default []
+            The parent namespace to list namespaces in.
+            Empty list represents root namespace.
+        page_token: str, optional
+            The token to use for pagination. If not present, start from the beginning.
+        limit: int, default 10
+            The size of the page to return.
+
+        Returns
+        -------
+        Iterable of str
+            List of immediate child namespace names
+        """
+        return []
+
+    def create_namespace(self, namespace: List[str]) -> None:
+        """Create a new namespace.
+
+        Parameters
+        ----------
+        namespace: List[str]
+            The namespace identifier to create.
+        """
+        raise NotImplementedError(
+            "Namespace operations are not supported for this connection type"
+        )
+
+    def drop_namespace(self, namespace: List[str]) -> None:
+        """Drop a namespace.
+
+        Parameters
+        ----------
+        namespace: List[str]
+            The namespace identifier to drop.
+        """
+        raise NotImplementedError(
+            "Namespace operations are not supported for this connection type"
+        )
+
     @abstractmethod
     def table_names(
-        self, page_token: Optional[str] = None, limit: int = 10
+        self,
+        page_token: Optional[str] = None,
+        limit: int = 10,
+        *,
+        namespace: List[str] = [],
     ) -> Iterable[str]:
         """List all tables in this database, in sorted order
 
         Parameters
         ----------
+        namespace: List[str], default []
+            The namespace to list tables in.
+            Empty list represents root namespace.
         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.
@@ -77,6 +133,7 @@ class DBConnection(EnforceOverrides):
         fill_value: float = 0.0,
         embedding_functions: Optional[List[EmbeddingFunctionConfig]] = None,
         *,
+        namespace: List[str] = [],
         storage_options: Optional[Dict[str, str]] = None,
         data_storage_version: Optional[str] = None,
         enable_v2_manifest_paths: Optional[bool] = None,
@@ -87,6 +144,9 @@ class DBConnection(EnforceOverrides):
         ----------
         name: str
             The name of the table.
+        namespace: List[str], default []
+            The namespace to create the table in.
+            Empty list represents root namespace.
         data: The data to initialize the table, *optional*
             User must provide at least one of `data` or `schema`.
             Acceptable types are:
@@ -238,6 +298,7 @@ class DBConnection(EnforceOverrides):
         self,
         name: str,
         *,
+        namespace: List[str] = [],
         storage_options: Optional[Dict[str, str]] = None,
         index_cache_size: Optional[int] = None,
     ) -> Table:
@@ -247,6 +308,9 @@ class DBConnection(EnforceOverrides):
         ----------
         name: str
             The name of the table.
+        namespace: List[str], optional
+            The namespace to open the table from.
+            None or empty list represents root namespace.
         index_cache_size: int, default 256
             **Deprecated**: Use session-level cache configuration instead.
             Create a Session with custom cache sizes and pass it to lancedb.connect().
@@ -272,17 +336,26 @@ class DBConnection(EnforceOverrides):
         """
         raise NotImplementedError
 
-    def drop_table(self, name: str):
+    def drop_table(self, name: str, namespace: List[str] = []):
         """Drop a table from the database.
 
         Parameters
         ----------
         name: str
             The name of the table.
+        namespace: List[str], default []
+            The namespace to drop the table from.
+            Empty list represents root namespace.
         """
         raise NotImplementedError
 
-    def rename_table(self, cur_name: str, new_name: str):
+    def rename_table(
+        self,
+        cur_name: str,
+        new_name: str,
+        cur_namespace: List[str] = [],
+        new_namespace: List[str] = [],
+    ):
         """Rename a table in the database.
 
         Parameters
@@ -291,6 +364,12 @@ class DBConnection(EnforceOverrides):
             The current name of the table.
         new_name: str
             The new name of the table.
+        cur_namespace: List[str], optional
+            The namespace of the current table.
+            None or empty list represents root namespace.
+        new_namespace: List[str], optional
+            The namespace to move the table to.
+            If not specified, defaults to the same as cur_namespace.
         """
         raise NotImplementedError
 
@@ -301,9 +380,15 @@ class DBConnection(EnforceOverrides):
         """
         raise NotImplementedError
 
-    def drop_all_tables(self):
+    def drop_all_tables(self, namespace: List[str] = []):
         """
         Drop all tables from the database
+
+        Parameters
+        ----------
+        namespace: List[str], optional
+            The namespace to drop all tables from.
+            None or empty list represents root namespace.
         """
         raise NotImplementedError
 
@@ -404,18 +489,87 @@ class LanceDBConnection(DBConnection):
         conn = AsyncConnection(await lancedb_connect(self.uri))
         return await conn.table_names(start_after=start_after, limit=limit)
 
+    @override
+    def list_namespaces(
+        self,
+        namespace: List[str] = [],
+        page_token: Optional[str] = None,
+        limit: int = 10,
+    ) -> Iterable[str]:
+        """List immediate child namespace names in the given namespace.
+
+        Parameters
+        ----------
+        namespace: List[str], optional
+            The parent namespace to list namespaces in.
+            None or empty list represents root namespace.
+        page_token: str, optional
+            The token to use for pagination. If not present, start from the beginning.
+        limit: int, default 10
+            The size of the page to return.
+
+        Returns
+        -------
+        Iterable of str
+            List of immediate child namespace names
+        """
+        return LOOP.run(
+            self._conn.list_namespaces(
+                namespace=namespace, page_token=page_token, limit=limit
+            )
+        )
+
+    @override
+    def create_namespace(self, namespace: List[str]) -> None:
+        """Create a new namespace.
+
+        Parameters
+        ----------
+        namespace: List[str]
+            The namespace identifier to create.
+        """
+        LOOP.run(self._conn.create_namespace(namespace=namespace))
+
+    @override
+    def drop_namespace(self, namespace: List[str]) -> None:
+        """Drop a namespace.
+
+        Parameters
+        ----------
+        namespace: List[str]
+            The namespace identifier to drop.
+        """
+        return LOOP.run(self._conn.drop_namespace(namespace=namespace))
+
     @override
     def table_names(
-        self, page_token: Optional[str] = None, limit: int = 10
+        self,
+        page_token: Optional[str] = None,
+        limit: int = 10,
+        *,
+        namespace: List[str] = [],
     ) -> Iterable[str]:
         """Get the names of all tables in the database. The names are sorted.
 
+        Parameters
+        ----------
+        namespace: List[str], optional
+            The namespace to list tables in.
+        page_token: str, optional
+            The token to use for pagination.
+        limit: int, default 10
+            The maximum number of tables to return.
+
         Returns
         -------
         Iterator of str.
             A list of table names.
         """
-        return LOOP.run(self._conn.table_names(start_after=page_token, limit=limit))
+        return LOOP.run(
+            self._conn.table_names(
+                namespace=namespace, start_after=page_token, limit=limit
+            )
+        )
 
     def __len__(self) -> int:
         return len(self.table_names())
@@ -435,12 +589,18 @@ class LanceDBConnection(DBConnection):
         fill_value: float = 0.0,
         embedding_functions: Optional[List[EmbeddingFunctionConfig]] = None,
         *,
+        namespace: List[str] = [],
         storage_options: Optional[Dict[str, str]] = None,
         data_storage_version: Optional[str] = None,
         enable_v2_manifest_paths: Optional[bool] = None,
     ) -> LanceTable:
         """Create a table in the database.
 
+        Parameters
+        ----------
+        namespace: List[str], optional
+            The namespace to create the table in.
+
         See
         ---
         DBConnection.create_table
@@ -459,6 +619,7 @@ class LanceDBConnection(DBConnection):
             on_bad_vectors=on_bad_vectors,
             fill_value=fill_value,
             embedding_functions=embedding_functions,
+            namespace=namespace,
             storage_options=storage_options,
         )
         return tbl
@@ -468,6 +629,7 @@ class LanceDBConnection(DBConnection):
         self,
         name: str,
         *,
+        namespace: List[str] = [],
         storage_options: Optional[Dict[str, str]] = None,
         index_cache_size: Optional[int] = None,
     ) -> LanceTable:
@@ -477,6 +639,8 @@ class LanceDBConnection(DBConnection):
         ----------
         name: str
             The name of the table.
+        namespace: List[str], optional
+            The namespace to open the table from.
 
         Returns
         -------
@@ -496,26 +660,122 @@ class LanceDBConnection(DBConnection):
         return LanceTable.open(
             self,
             name,
+            namespace=namespace,
             storage_options=storage_options,
             index_cache_size=index_cache_size,
         )
 
+    def clone_table(
+        self,
+        target_table_name: str,
+        source_uri: str,
+        *,
+        target_namespace: List[str] = [],
+        source_version: Optional[int] = None,
+        source_tag: Optional[str] = None,
+        is_shallow: bool = True,
+    ) -> LanceTable:
+        """Clone a table from a source table.
+
+        A shallow clone creates a new table that shares the underlying data files
+        with the source table but has its own independent manifest. This allows
+        both the source and cloned tables to evolve independently while initially
+        sharing the same data, deletion, and index files.
+
+        Parameters
+        ----------
+        target_table_name: str
+            The name of the target table to create.
+        source_uri: str
+            The URI of the source table to clone from.
+        target_namespace: List[str], optional
+            The namespace for the target table.
+            None or empty list represents root namespace.
+        source_version: int, optional
+            The version of the source table to clone.
+        source_tag: str, optional
+            The tag of the source table to clone.
+        is_shallow: bool, default True
+            Whether to perform a shallow clone (True) or deep clone (False).
+            Currently only shallow clone is supported.
+
+        Returns
+        -------
+        A LanceTable object representing the cloned table.
+        """
+        LOOP.run(
+            self._conn.clone_table(
+                target_table_name,
+                source_uri,
+                target_namespace=target_namespace,
+                source_version=source_version,
+                source_tag=source_tag,
+                is_shallow=is_shallow,
+            )
+        )
+        return LanceTable.open(
+            self,
+            target_table_name,
+            namespace=target_namespace,
+        )
+
     @override
-    def drop_table(self, name: str, ignore_missing: bool = False):
+    def drop_table(
+        self,
+        name: str,
+        namespace: List[str] = [],
+        ignore_missing: bool = False,
+    ):
         """Drop a table from the database.
 
         Parameters
         ----------
         name: str
             The name of the table.
+        namespace: List[str], optional
+            The namespace to drop the table from.
         ignore_missing: bool, default False
             If True, ignore if the table does not exist.
         """
-        LOOP.run(self._conn.drop_table(name, ignore_missing=ignore_missing))
+        LOOP.run(
+            self._conn.drop_table(
+                name, namespace=namespace, ignore_missing=ignore_missing
+            )
+        )
 
     @override
-    def drop_all_tables(self):
-        LOOP.run(self._conn.drop_all_tables())
+    def drop_all_tables(self, namespace: List[str] = []):
+        LOOP.run(self._conn.drop_all_tables(namespace=namespace))
+
+    @override
+    def rename_table(
+        self,
+        cur_name: str,
+        new_name: str,
+        cur_namespace: List[str] = [],
+        new_namespace: List[str] = [],
+    ):
+        """Rename a table in the database.
+
+        Parameters
+        ----------
+        cur_name: str
+            The current name of the table.
+        new_name: str
+            The new name of the table.
+        cur_namespace: List[str], optional
+            The namespace of the current table.
+        new_namespace: List[str], optional
+            The namespace to move the table to.
+        """
+        LOOP.run(
+            self._conn.rename_table(
+                cur_name,
+                new_name,
+                cur_namespace=cur_namespace,
+                new_namespace=new_namespace,
+            )
+        )
 
     @deprecation.deprecated(
         deprecated_in="0.15.1",
@@ -588,13 +848,67 @@ class AsyncConnection(object):
     def uri(self) -> str:
         return self._inner.uri
 
+    async def list_namespaces(
+        self,
+        namespace: List[str] = [],
+        page_token: Optional[str] = None,
+        limit: int = 10,
+    ) -> Iterable[str]:
+        """List immediate child namespace names in the given namespace.
+
+        Parameters
+        ----------
+        namespace: List[str], optional
+            The parent namespace to list namespaces in.
+            None or empty list represents root namespace.
+        page_token: str, optional
+            The token to use for pagination. If not present, start from the beginning.
+        limit: int, default 10
+            The size of the page to return.
+
+        Returns
+        -------
+        Iterable of str
+            List of immediate child namespace names (not full paths)
+        """
+        return await self._inner.list_namespaces(
+            namespace=namespace, page_token=page_token, limit=limit
+        )
+
+    async def create_namespace(self, namespace: List[str]) -> None:
+        """Create a new namespace.
+
+        Parameters
+        ----------
+        namespace: List[str]
+            The namespace identifier to create.
+        """
+        await self._inner.create_namespace(namespace)
+
+    async def drop_namespace(self, namespace: List[str]) -> None:
+        """Drop a namespace.
+
+        Parameters
+        ----------
+        namespace: List[str]
+            The namespace identifier to drop.
+        """
+        await self._inner.drop_namespace(namespace)
+
     async def table_names(
-        self, *, start_after: Optional[str] = None, limit: Optional[int] = None
+        self,
+        *,
+        namespace: List[str] = [],
+        start_after: Optional[str] = None,
+        limit: Optional[int] = None,
     ) -> Iterable[str]:
         """List all tables in this database, in sorted order
 
         Parameters
         ----------
+        namespace: List[str], optional
+            The namespace to list tables in.
+            None or empty list represents root namespace.
         start_after: str, optional
             If present, only return names that come lexicographically after the supplied
             value.
@@ -608,7 +922,9 @@ class AsyncConnection(object):
         -------
         Iterable of str
         """
-        return await self._inner.table_names(start_after=start_after, limit=limit)
+        return await self._inner.table_names(
+            namespace=namespace, start_after=start_after, limit=limit
+        )
 
     async def create_table(
         self,
@@ -621,6 +937,7 @@ class AsyncConnection(object):
         fill_value: Optional[float] = None,
         storage_options: Optional[Dict[str, str]] = None,
         *,
+        namespace: List[str] = [],
         embedding_functions: Optional[List[EmbeddingFunctionConfig]] = None,
     ) -> AsyncTable:
         """Create an [AsyncTable][lancedb.table.AsyncTable] in the database.
@@ -629,6 +946,9 @@ class AsyncConnection(object):
         ----------
         name: str
             The name of the table.
+        namespace: List[str], default []
+            The namespace to create the table in.
+            Empty list represents root namespace.
         data: The data to initialize the table, *optional*
             User must provide at least one of `data` or `schema`.
             Acceptable types are:
@@ -807,6 +1127,7 @@ class AsyncConnection(object):
                 name,
                 mode,
                 schema,
+                namespace=namespace,
                 storage_options=storage_options,
             )
         else:
@@ -815,6 +1136,7 @@ class AsyncConnection(object):
                 name,
                 mode,
                 data,
+                namespace=namespace,
                 storage_options=storage_options,
             )
 
@@ -823,6 +1145,8 @@ class AsyncConnection(object):
     async def open_table(
         self,
         name: str,
+        *,
+        namespace: List[str] = [],
         storage_options: Optional[Dict[str, str]] = None,
         index_cache_size: Optional[int] = None,
     ) -> AsyncTable:
@@ -832,6 +1156,9 @@ class AsyncConnection(object):
         ----------
         name: str
             The name of the table.
+        namespace: List[str], optional
+            The namespace to open the table from.
+            None or empty list represents root namespace.
         storage_options: dict, optional
             Additional options for the storage backend. Options already set on the
             connection will be inherited by the table, but can be overridden here.
@@ -855,42 +1182,125 @@ class AsyncConnection(object):
         -------
         A LanceTable object representing the table.
         """
-        table = await self._inner.open_table(name, storage_options, index_cache_size)
+        table = await self._inner.open_table(
+            name,
+            namespace=namespace,
+            storage_options=storage_options,
+            index_cache_size=index_cache_size,
+        )
         return AsyncTable(table)
 
-    async def rename_table(self, old_name: str, new_name: str):
+    async def clone_table(
+        self,
+        target_table_name: str,
+        source_uri: str,
+        *,
+        target_namespace: List[str] = [],
+        source_version: Optional[int] = None,
+        source_tag: Optional[str] = None,
+        is_shallow: bool = True,
+    ) -> AsyncTable:
+        """Clone a table from a source table.
+
+        A shallow clone creates a new table that shares the underlying data files
+        with the source table but has its own independent manifest. This allows
+        both the source and cloned tables to evolve independently while initially
+        sharing the same data, deletion, and index files.
+
+        Parameters
+        ----------
+        target_table_name: str
+            The name of the target table to create.
+        source_uri: str
+            The URI of the source table to clone from.
+        target_namespace: List[str], optional
+            The namespace for the target table.
+            None or empty list represents root namespace.
+        source_version: int, optional
+            The version of the source table to clone.
+        source_tag: str, optional
+            The tag of the source table to clone.
+        is_shallow: bool, default True
+            Whether to perform a shallow clone (True) or deep clone (False).
+            Currently only shallow clone is supported.
+
+        Returns
+        -------
+        An AsyncTable object representing the cloned table.
+        """
+        table = await self._inner.clone_table(
+            target_table_name,
+            source_uri,
+            target_namespace=target_namespace,
+            source_version=source_version,
+            source_tag=source_tag,
+            is_shallow=is_shallow,
+        )
+        return AsyncTable(table)
+
+    async def rename_table(
+        self,
+        cur_name: str,
+        new_name: str,
+        cur_namespace: List[str] = [],
+        new_namespace: List[str] = [],
+    ):
         """Rename a table in the database.
 
         Parameters
         ----------
-        old_name: str
+        cur_name: str
             The current name of the table.
         new_name: str
             The new name of the table.
+        cur_namespace: List[str], optional
+            The namespace of the current table.
+            None or empty list represents root namespace.
+        new_namespace: List[str], optional
+            The namespace to move the table to.
+            If not specified, defaults to the same as cur_namespace.
         """
-        await self._inner.rename_table(old_name, new_name)
+        await self._inner.rename_table(
+            cur_name, new_name, cur_namespace=cur_namespace, new_namespace=new_namespace
+        )
 
-    async def drop_table(self, name: str, *, ignore_missing: bool = False):
+    async def drop_table(
+        self,
+        name: str,
+        *,
+        namespace: List[str] = [],
+        ignore_missing: bool = False,
+    ):
         """Drop a table from the database.
 
         Parameters
         ----------
         name: str
             The name of the table.
+        namespace: List[str], default []
+            The namespace to drop the table from.
+            Empty list represents root namespace.
         ignore_missing: bool, default False
             If True, ignore if the table does not exist.
         """
         try:
-            await self._inner.drop_table(name)
+            await self._inner.drop_table(name, namespace=namespace)
         except ValueError as e:
             if not ignore_missing:
                 raise e
             if f"Table '{name}' was not found" not in str(e):
                 raise e
 
-    async def drop_all_tables(self):
-        """Drop all tables from the database."""
-        await self._inner.drop_all_tables()
+    async def drop_all_tables(self, namespace: List[str] = []):
+        """Drop all tables from the database.
+
+        Parameters
+        ----------
+        namespace: List[str], optional
+            The namespace to drop all tables from.
+            None or empty list represents root namespace.
+        """
+        await self._inner.drop_all_tables(namespace=namespace)
 
     @deprecation.deprecated(
         deprecated_in="0.15.1",

python/python/lancedb/embeddings/registry.py

@@ -122,7 +122,7 @@ class EmbeddingFunctionRegistry:
             obj["vector_column"]: EmbeddingFunctionConfig(
                 vector_column=obj["vector_column"],
                 source_column=obj["source_column"],
-                function=self.get(obj["name"])(**obj["model"]),
+                function=self.get(obj["name"]).create(**obj["model"]),
             )
             for obj in raw_list
         }

python/python/lancedb/index.py

@@ -251,6 +251,13 @@ class HnswPq:
         results. In most cases, there is no benefit to setting this higher than 500.
         This value should be set to a value that is not less than `ef` in the
         search phase.
+
+    target_partition_size, default is 1,048,576
+
+        The target size of each partition.
+
+        This value controls the tradeoff between search performance and accuracy.
+        faster search but less accurate results as higher value.
     """
 
     distance_type: Literal["l2", "cosine", "dot"] = "l2"
@@ -261,6 +268,7 @@ class HnswPq:
     sample_rate: int = 256
     m: int = 20
     ef_construction: int = 300
+    target_partition_size: Optional[int] = None
 
 
 @dataclass
@@ -351,6 +359,12 @@ class HnswSq:
         This value should be set to a value that is not less than `ef` in the search
         phase.
 
+    target_partition_size, default is 1,048,576
+
+        The target size of each partition.
+
+        This value controls the tradeoff between search performance and accuracy.
+        faster search but less accurate results as higher value.
     """
 
     distance_type: Literal["l2", "cosine", "dot"] = "l2"
@@ -359,6 +373,7 @@ class HnswSq:
     sample_rate: int = 256
     m: int = 20
     ef_construction: int = 300
+    target_partition_size: Optional[int] = None
 
 
 @dataclass
@@ -444,12 +459,20 @@ class IvfFlat:
         cases the default should be sufficient.
 
         The default value is 256.
+
+    target_partition_size, default is 8192
+
+        The target size of each partition.
+
+        This value controls the tradeoff between search performance and accuracy.
+        faster search but less accurate results as higher value.
     """
 
     distance_type: Literal["l2", "cosine", "dot", "hamming"] = "l2"
     num_partitions: Optional[int] = None
     max_iterations: int = 50
     sample_rate: int = 256
+    target_partition_size: Optional[int] = None
 
 
 @dataclass
@@ -564,6 +587,13 @@ class IvfPq:
         cases the default should be sufficient.
 
         The default value is 256.
+
+    target_partition_size, default is 8192
+
+        The target size of each partition.
+
+        This value controls the tradeoff between search performance and accuracy.
+        faster search but less accurate results as higher value.
     """
 
     distance_type: Literal["l2", "cosine", "dot"] = "l2"
@@ -572,6 +602,7 @@ class IvfPq:
     num_bits: int = 8
     max_iterations: int = 50
     sample_rate: int = 256
+    target_partition_size: Optional[int] = None
 
 
 __all__ = [

python/python/lancedb/namespace.py

@@ -0,0 +1,401 @@
+# SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright The LanceDB Authors
+
+"""
+LanceDB Namespace integration module.
+
+This module provides integration with lance_namespace for managing tables
+through a namespace abstraction.
+"""
+
+from __future__ import annotations
+
+from typing import Dict, Iterable, List, Optional, Union
+import os
+
+from lancedb.db import DBConnection
+from lancedb.table import LanceTable, Table
+from lancedb.util import validate_table_name
+from lancedb.common import validate_schema
+from lancedb.table import sanitize_create_table
+from overrides import override
+
+from lance_namespace import LanceNamespace, connect as namespace_connect
+from lance_namespace_urllib3_client.models import (
+    ListTablesRequest,
+    DescribeTableRequest,
+    CreateTableRequest,
+    DropTableRequest,
+    ListNamespacesRequest,
+    CreateNamespaceRequest,
+    DropNamespaceRequest,
+    JsonArrowSchema,
+    JsonArrowField,
+    JsonArrowDataType,
+)
+
+import pyarrow as pa
+from datetime import timedelta
+from lancedb.pydantic import LanceModel
+from lancedb.common import DATA
+from lancedb.embeddings import EmbeddingFunctionConfig
+from ._lancedb import Session
+
+
+def _convert_pyarrow_type_to_json(arrow_type: pa.DataType) -> JsonArrowDataType:
+    """Convert PyArrow DataType to JsonArrowDataType."""
+    if pa.types.is_null(arrow_type):
+        type_name = "null"
+    elif pa.types.is_boolean(arrow_type):
+        type_name = "bool"
+    elif pa.types.is_int8(arrow_type):
+        type_name = "int8"
+    elif pa.types.is_uint8(arrow_type):
+        type_name = "uint8"
+    elif pa.types.is_int16(arrow_type):
+        type_name = "int16"
+    elif pa.types.is_uint16(arrow_type):
+        type_name = "uint16"
+    elif pa.types.is_int32(arrow_type):
+        type_name = "int32"
+    elif pa.types.is_uint32(arrow_type):
+        type_name = "uint32"
+    elif pa.types.is_int64(arrow_type):
+        type_name = "int64"
+    elif pa.types.is_uint64(arrow_type):
+        type_name = "uint64"
+    elif pa.types.is_float32(arrow_type):
+        type_name = "float32"
+    elif pa.types.is_float64(arrow_type):
+        type_name = "float64"
+    elif pa.types.is_string(arrow_type):
+        type_name = "utf8"
+    elif pa.types.is_binary(arrow_type):
+        type_name = "binary"
+    elif pa.types.is_list(arrow_type):
+        # For list types, we need more complex handling
+        type_name = "list"
+    elif pa.types.is_fixed_size_list(arrow_type):
+        type_name = "fixed_size_list"
+    else:
+        # Default to string representation for unsupported types
+        type_name = str(arrow_type)
+
+    return JsonArrowDataType(type=type_name)
+
+
+def _convert_pyarrow_schema_to_json(schema: pa.Schema) -> JsonArrowSchema:
+    """Convert PyArrow Schema to JsonArrowSchema."""
+    fields = []
+    for field in schema:
+        json_field = JsonArrowField(
+            name=field.name,
+            type=_convert_pyarrow_type_to_json(field.type),
+            nullable=field.nullable,
+            metadata=field.metadata,
+        )
+        fields.append(json_field)
+
+    return JsonArrowSchema(fields=fields, metadata=schema.metadata)
+
+
+class LanceNamespaceDBConnection(DBConnection):
+    """
+    A LanceDB connection that uses a namespace for table management.
+
+    This connection delegates table URI resolution to a lance_namespace instance,
+    while using the standard LanceTable for actual table operations.
+    """
+
+    def __init__(
+        self,
+        namespace: LanceNamespace,
+        *,
+        read_consistency_interval: Optional[timedelta] = None,
+        storage_options: Optional[Dict[str, str]] = None,
+        session: Optional[Session] = None,
+    ):
+        """
+        Initialize a namespace-based LanceDB connection.
+
+        Parameters
+        ----------
+        namespace : LanceNamespace
+            The namespace instance to use for table management
+        read_consistency_interval : Optional[timedelta]
+            The interval at which to check for updates to the table from other
+            processes. If None, then consistency is not checked.
+        storage_options : Optional[Dict[str, str]]
+            Additional options for the storage backend
+        session : Optional[Session]
+            A session to use for this connection
+        """
+        self._ns = namespace
+        self.read_consistency_interval = read_consistency_interval
+        self.storage_options = storage_options or {}
+        self.session = session
+
+    @override
+    def table_names(
+        self,
+        page_token: Optional[str] = None,
+        limit: int = 10,
+        *,
+        namespace: List[str] = [],
+    ) -> Iterable[str]:
+        request = ListTablesRequest(id=namespace, page_token=page_token, limit=limit)
+        response = self._ns.list_tables(request)
+        return response.tables if response.tables else []
+
+    @override
+    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,
+        *,
+        namespace: List[str] = [],
+        storage_options: Optional[Dict[str, str]] = None,
+        data_storage_version: Optional[str] = None,
+        enable_v2_manifest_paths: Optional[bool] = None,
+    ) -> Table:
+        if mode.lower() not in ["create", "overwrite"]:
+            raise ValueError("mode must be either 'create' or 'overwrite'")
+        validate_table_name(name)
+
+        # TODO: support passing data
+        if data is not None:
+            raise ValueError(
+                "create_table currently only supports creating empty tables (data=None)"
+            )
+
+        # Prepare schema
+        metadata = None
+        if embedding_functions is not None:
+            from lancedb.embeddings.registry import EmbeddingFunctionRegistry
+
+            registry = EmbeddingFunctionRegistry.get_instance()
+            metadata = registry.get_table_metadata(embedding_functions)
+
+        data, schema = sanitize_create_table(
+            data, schema, metadata, on_bad_vectors, fill_value
+        )
+        validate_schema(schema)
+
+        # Convert PyArrow schema to JsonArrowSchema
+        json_schema = _convert_pyarrow_schema_to_json(schema)
+
+        # Create table request with namespace
+        table_id = namespace + [name]
+        request = CreateTableRequest(id=table_id, var_schema=json_schema)
+
+        # Create empty Arrow IPC stream bytes
+        import pyarrow.ipc as ipc
+        import io
+
+        empty_table = pa.Table.from_arrays(
+            [pa.array([], type=field.type) for field in schema], schema=schema
+        )
+        buffer = io.BytesIO()
+        with ipc.new_stream(buffer, schema) as writer:
+            writer.write_table(empty_table)
+        request_data = buffer.getvalue()
+
+        self._ns.create_table(request, request_data)
+        return self.open_table(
+            name, namespace=namespace, storage_options=storage_options
+        )
+
+    @override
+    def open_table(
+        self,
+        name: str,
+        *,
+        namespace: List[str] = [],
+        storage_options: Optional[Dict[str, str]] = None,
+        index_cache_size: Optional[int] = None,
+    ) -> Table:
+        table_id = namespace + [name]
+        request = DescribeTableRequest(id=table_id)
+        response = self._ns.describe_table(request)
+
+        merged_storage_options = dict()
+        if storage_options:
+            merged_storage_options.update(storage_options)
+        if response.storage_options:
+            merged_storage_options.update(response.storage_options)
+
+        return self._lance_table_from_uri(
+            response.location,
+            storage_options=merged_storage_options,
+            index_cache_size=index_cache_size,
+        )
+
+    @override
+    def drop_table(self, name: str, namespace: List[str] = []):
+        # Use namespace drop_table directly
+        table_id = namespace + [name]
+        request = DropTableRequest(id=table_id)
+        self._ns.drop_table(request)
+
+    @override
+    def rename_table(
+        self,
+        cur_name: str,
+        new_name: str,
+        cur_namespace: List[str] = [],
+        new_namespace: List[str] = [],
+    ):
+        raise NotImplementedError(
+            "rename_table is not supported for namespace connections"
+        )
+
+    @override
+    def drop_database(self):
+        raise NotImplementedError(
+            "drop_database is deprecated, use drop_all_tables instead"
+        )
+
+    @override
+    def drop_all_tables(self, namespace: List[str] = []):
+        for table_name in self.table_names(namespace=namespace):
+            self.drop_table(table_name, namespace=namespace)
+
+    @override
+    def list_namespaces(
+        self,
+        namespace: List[str] = [],
+        page_token: Optional[str] = None,
+        limit: int = 10,
+    ) -> Iterable[str]:
+        """
+        List child namespaces under the given namespace.
+
+        Parameters
+        ----------
+        namespace : Optional[List[str]]
+            The parent namespace to list children from.
+            If None, lists root-level namespaces.
+        page_token : Optional[str]
+            Pagination token for listing results.
+        limit : int
+            Maximum number of namespaces to return.
+
+        Returns
+        -------
+        Iterable[str]
+            Names of child namespaces.
+        """
+        request = ListNamespacesRequest(
+            id=namespace, page_token=page_token, limit=limit
+        )
+        response = self._ns.list_namespaces(request)
+        return response.namespaces if response.namespaces else []
+
+    @override
+    def create_namespace(self, namespace: List[str]) -> None:
+        """
+        Create a new namespace.
+
+        Parameters
+        ----------
+        namespace : List[str]
+            The namespace path to create.
+        """
+        request = CreateNamespaceRequest(id=namespace)
+        self._ns.create_namespace(request)
+
+    @override
+    def drop_namespace(self, namespace: List[str]) -> None:
+        """
+        Drop a namespace.
+
+        Parameters
+        ----------
+        namespace : List[str]
+            The namespace path to drop.
+        """
+        request = DropNamespaceRequest(id=namespace)
+        self._ns.drop_namespace(request)
+
+    def _lance_table_from_uri(
+        self,
+        table_uri: str,
+        *,
+        storage_options: Optional[Dict[str, str]] = None,
+        index_cache_size: Optional[int] = None,
+    ) -> LanceTable:
+        # Extract the base path and table name from the URI
+        if table_uri.endswith(".lance"):
+            base_path = os.path.dirname(table_uri)
+            table_name = os.path.basename(table_uri)[:-6]  # Remove .lance
+        else:
+            raise ValueError(f"Invalid table URI: {table_uri}")
+
+        from lancedb.db import LanceDBConnection
+
+        temp_conn = LanceDBConnection(
+            base_path,
+            read_consistency_interval=self.read_consistency_interval,
+            storage_options={**self.storage_options, **(storage_options or {})},
+            session=self.session,
+        )
+
+        # Open the table using the temporary connection
+        return LanceTable.open(
+            temp_conn,
+            table_name,
+            storage_options=storage_options,
+            index_cache_size=index_cache_size,
+        )
+
+
+def connect_namespace(
+    impl: str,
+    properties: Dict[str, str],
+    *,
+    read_consistency_interval: Optional[timedelta] = None,
+    storage_options: Optional[Dict[str, str]] = None,
+    session: Optional[Session] = None,
+) -> LanceNamespaceDBConnection:
+    """
+    Connect to a LanceDB database through a namespace.
+
+    Parameters
+    ----------
+    impl : str
+        The namespace implementation to use. For examples:
+        - "dir" for DirectoryNamespace
+        - "rest" for REST-based namespace
+        - Full module path for custom implementations
+    properties : Dict[str, str]
+        Configuration properties for the namespace implementation.
+        Different namespace implemenation has different config properties.
+        For example, use DirectoryNamespace with {"root": "/path/to/directory"}
+    read_consistency_interval : Optional[timedelta]
+        The interval at which to check for updates to the table from other
+        processes. If None, then consistency is not checked.
+    storage_options : Optional[Dict[str, str]]
+        Additional options for the storage backend
+    session : Optional[Session]
+        A session to use for this connection
+
+    Returns
+    -------
+    LanceNamespaceDBConnection
+        A namespace-based connection to LanceDB
+    """
+    namespace = namespace_connect(impl, properties)
+
+    # Return the namespace-based connection
+    return LanceNamespaceDBConnection(
+        namespace,
+        read_consistency_interval=read_consistency_interval,
+        storage_options=storage_options,
+        session=session,
+    )

python/python/lancedb/query.py

@@ -28,6 +28,7 @@ import pyarrow.fs as pa_fs
 import pydantic
 
 from lancedb.pydantic import PYDANTIC_VERSION
+from lancedb.background_loop import LOOP
 
 from . import __version__
 from .arrow import AsyncRecordBatchReader
@@ -48,6 +49,7 @@ if TYPE_CHECKING:
     from ._lancedb import FTSQuery as LanceFTSQuery
     from ._lancedb import HybridQuery as LanceHybridQuery
     from ._lancedb import VectorQuery as LanceVectorQuery
+    from ._lancedb import TakeQuery as LanceTakeQuery
     from ._lancedb import PyQueryRequest
     from .common import VEC
     from .pydantic import LanceModel
@@ -910,7 +912,7 @@ class LanceQueryBuilder(ABC):
         ProjectionExec: expr=[vector@0 as vector, _distance@2 as _distance]
           GlobalLimitExec: skip=0, fetch=10
             FilterExec: _distance@2 IS NOT NULL
-              SortExec: TopK(fetch=10), expr=[_distance@2 ASC NULLS LAST], preserve_partitioning=[false]
+              SortExec: TopK(fetch=10), expr=[_distance@2 ASC NULLS LAST, _rowid@1 ASC NULLS LAST], preserve_partitioning=[false]
                 KNNVectorDistance: metric=l2
                   LanceRead: uri=..., projection=[vector], ...
 
@@ -941,20 +943,22 @@ class LanceQueryBuilder(ABC):
         >>> query = [100, 100]
         >>> plan = table.search(query).analyze_plan()
         >>> print(plan)  # doctest: +ELLIPSIS, +NORMALIZE_WHITESPACE
-        AnalyzeExec verbose=true, metrics=[]
-          TracedExec, metrics=[]
-            ProjectionExec: expr=[...], metrics=[...]
-              GlobalLimitExec: skip=0, fetch=10, metrics=[...]
+        AnalyzeExec verbose=true, metrics=[], cumulative_cpu=...
+          TracedExec, metrics=[], cumulative_cpu=...
+            ProjectionExec: expr=[...], metrics=[...], cumulative_cpu=...
+              GlobalLimitExec: skip=0, fetch=10, metrics=[...], cumulative_cpu=...
                 FilterExec: _distance@2 IS NOT NULL,
-                metrics=[output_rows=..., elapsed_compute=...]
+                metrics=[output_rows=..., elapsed_compute=...], cumulative_cpu=...
                   SortExec: TopK(fetch=10), expr=[...],
                   preserve_partitioning=[...],
-                  metrics=[output_rows=..., elapsed_compute=..., row_replacements=...]
+                  metrics=[output_rows=..., elapsed_compute=..., row_replacements=...],
+                  cumulative_cpu=...
                     KNNVectorDistance: metric=l2,
-                    metrics=[output_rows=..., elapsed_compute=..., output_batches=...]
+                    metrics=[output_rows=..., elapsed_compute=..., output_batches=...],
+                    cumulative_cpu=...
                       LanceRead: uri=..., projection=[vector], ...
                       metrics=[output_rows=..., elapsed_compute=...,
-                      bytes_read=..., iops=..., requests=...]
+                      bytes_read=..., iops=..., requests=...], cumulative_cpu=...
 
         Returns
         -------
@@ -2041,11 +2045,11 @@ class LanceHybridQueryBuilder(LanceQueryBuilder):
         >>> plan = table.search(query).explain_plan(True)
         >>> print(plan) # doctest: +ELLIPSIS, +NORMALIZE_WHITESPACE
         ProjectionExec: expr=[vector@0 as vector, _distance@2 as _distance]
-        GlobalLimitExec: skip=0, fetch=10
-          FilterExec: _distance@2 IS NOT NULL
-            SortExec: TopK(fetch=10), expr=[_distance@2 ASC NULLS LAST], preserve_partitioning=[false]
-              KNNVectorDistance: metric=l2
-                LanceRead: uri=..., projection=[vector], ...
+          GlobalLimitExec: skip=0, fetch=10
+            FilterExec: _distance@2 IS NOT NULL
+              SortExec: TopK(fetch=10), expr=[_distance@2 ASC NULLS LAST, _rowid@1 ASC NULLS LAST], preserve_partitioning=[false]
+                KNNVectorDistance: metric=l2
+                  LanceRead: uri=..., projection=[vector], ...
 
         Parameters
         ----------
@@ -2139,7 +2143,11 @@ class LanceHybridQueryBuilder(LanceQueryBuilder):
 
 
 class AsyncQueryBase(object):
-    def __init__(self, inner: Union[LanceQuery, LanceVectorQuery]):
+    """
+    Base class for all async queries (take, scan, vector, fts, hybrid)
+    """
+
+    def __init__(self, inner: Union[LanceQuery, LanceVectorQuery, LanceTakeQuery]):
         """
         Construct an AsyncQueryBase
 
@@ -2149,27 +2157,14 @@ class AsyncQueryBase(object):
         self._inner = inner
 
     def to_query_object(self) -> Query:
+        """
+        Convert the query into a query object
+
+        This is currently experimental but can be useful as the query object is pure
+        python and more easily serializable.
+        """
         return Query.from_inner(self._inner.to_query_request())
 
-    def where(self, predicate: str) -> Self:
-        """
-        Only return rows matching the given predicate
-
-        The predicate should be supplied as an SQL query string.
-
-        Examples
-        --------
-
-        >>> predicate = "x > 10"
-        >>> predicate = "y > 0 AND y < 100"
-        >>> predicate = "x > 5 OR y = 'test'"
-
-        Filtering performance can often be improved by creating a scalar index
-        on the filter column(s).
-        """
-        self._inner.where(predicate)
-        return self
-
     def select(self, columns: Union[List[str], dict[str, str]]) -> Self:
         """
         Return only the specified columns.
@@ -2208,42 +2203,6 @@ class AsyncQueryBase(object):
             raise TypeError("columns must be a list of column names or a dict")
         return self
 
-    def limit(self, limit: int) -> Self:
-        """
-        Set the maximum number of results to return.
-
-        By default, a plain search has no limit.  If this method is not
-        called then every valid row from the table will be returned.
-        """
-        self._inner.limit(limit)
-        return self
-
-    def offset(self, offset: int) -> Self:
-        """
-        Set the offset for the results.
-
-        Parameters
-        ----------
-        offset: int
-            The offset to start fetching results from.
-        """
-        self._inner.offset(offset)
-        return self
-
-    def fast_search(self) -> Self:
-        """
-        Skip searching un-indexed data.
-
-        This can make queries faster, but will miss any data that has not been
-        indexed.
-
-        !!! tip
-            You can add new data into an existing index by calling
-            [AsyncTable.optimize][lancedb.table.AsyncTable.optimize].
-        """
-        self._inner.fast_search()
-        return self
-
     def with_row_id(self) -> Self:
         """
         Include the _rowid column in the results.
@@ -2251,27 +2210,6 @@ class AsyncQueryBase(object):
         self._inner.with_row_id()
         return self
 
-    def postfilter(self) -> Self:
-        """
-        If this is called then filtering will happen after the search instead of
-        before.
-        By default filtering will be performed before the search.  This is how
-        filtering is typically understood to work.  This prefilter step does add some
-        additional latency.  Creating a scalar index on the filter column(s) can
-        often improve this latency.  However, sometimes a filter is too complex or
-        scalar indices cannot be applied to the column.  In these cases postfiltering
-        can be used instead of prefiltering to improve latency.
-        Post filtering applies the filter to the results of the search.  This
-        means we only run the filter on a much smaller set of data.  However, it can
-        cause the query to return fewer than `limit` results (or even no results) if
-        none of the nearest results match the filter.
-        Post filtering happens during the "refine stage" (described in more detail in
-        @see {@link VectorQuery#refineFactor}).  This means that setting a higher refine
-        factor can often help restore some of the results lost by post filtering.
-        """
-        self._inner.postfilter()
-        return self
-
     async def to_batches(
         self,
         *,
@@ -2295,7 +2233,9 @@ class AsyncQueryBase(object):
             complete within the specified time, an error will be raised.
         """
         return AsyncRecordBatchReader(
-            await self._inner.execute(max_batch_length, timeout)
+            await self._inner.execute(
+                max_batch_length=max_batch_length, timeout=timeout
+            )
         )
 
     async def to_arrow(self, timeout: Optional[timedelta] = None) -> pa.Table:
@@ -2429,7 +2369,7 @@ class AsyncQueryBase(object):
         ProjectionExec: expr=[vector@0 as vector, _distance@2 as _distance]
           GlobalLimitExec: skip=0, fetch=10
             FilterExec: _distance@2 IS NOT NULL
-              SortExec: TopK(fetch=10), expr=[_distance@2 ASC NULLS LAST], preserve_partitioning=[false]
+              SortExec: TopK(fetch=10), expr=[_distance@2 ASC NULLS LAST, _rowid@1 ASC NULLS LAST], preserve_partitioning=[false]
                 KNNVectorDistance: metric=l2
                   LanceRead: uri=..., projection=[vector], ...
 
@@ -2454,7 +2394,98 @@ class AsyncQueryBase(object):
         return await self._inner.analyze_plan()
 
 
-class AsyncQuery(AsyncQueryBase):
+class AsyncStandardQuery(AsyncQueryBase):
+    """
+    Base class for "standard" async queries (all but take currently)
+    """
+
+    def __init__(self, inner: Union[LanceQuery, LanceVectorQuery]):
+        """
+        Construct an AsyncStandardQuery
+
+        This method is not intended to be called directly.  Instead, use the
+        [AsyncTable.query][lancedb.table.AsyncTable.query] method to create a query.
+        """
+        super().__init__(inner)
+
+    def where(self, predicate: str) -> Self:
+        """
+        Only return rows matching the given predicate
+
+        The predicate should be supplied as an SQL query string.
+
+        Examples
+        --------
+
+        >>> predicate = "x > 10"
+        >>> predicate = "y > 0 AND y < 100"
+        >>> predicate = "x > 5 OR y = 'test'"
+
+        Filtering performance can often be improved by creating a scalar index
+        on the filter column(s).
+        """
+        self._inner.where(predicate)
+        return self
+
+    def limit(self, limit: int) -> Self:
+        """
+        Set the maximum number of results to return.
+
+        By default, a plain search has no limit.  If this method is not
+        called then every valid row from the table will be returned.
+        """
+        self._inner.limit(limit)
+        return self
+
+    def offset(self, offset: int) -> Self:
+        """
+        Set the offset for the results.
+
+        Parameters
+        ----------
+        offset: int
+            The offset to start fetching results from.
+        """
+        self._inner.offset(offset)
+        return self
+
+    def fast_search(self) -> Self:
+        """
+        Skip searching un-indexed data.
+
+        This can make queries faster, but will miss any data that has not been
+        indexed.
+
+        !!! tip
+            You can add new data into an existing index by calling
+            [AsyncTable.optimize][lancedb.table.AsyncTable.optimize].
+        """
+        self._inner.fast_search()
+        return self
+
+    def postfilter(self) -> Self:
+        """
+        If this is called then filtering will happen after the search instead of
+        before.
+        By default filtering will be performed before the search.  This is how
+        filtering is typically understood to work.  This prefilter step does add some
+        additional latency.  Creating a scalar index on the filter column(s) can
+        often improve this latency.  However, sometimes a filter is too complex or
+        scalar indices cannot be applied to the column.  In these cases postfiltering
+        can be used instead of prefiltering to improve latency.
+        Post filtering applies the filter to the results of the search.  This
+        means we only run the filter on a much smaller set of data.  However, it can
+        cause the query to return fewer than `limit` results (or even no results) if
+        none of the nearest results match the filter.
+        Post filtering happens during the "refine stage" (described in more detail in
+        @see {@link VectorQuery#refineFactor}).  This means that setting a higher refine
+        factor can often help restore some of the results lost by post filtering.
+        """
+        self._inner.postfilter()
+        return self
+
+
+class AsyncQuery(AsyncStandardQuery):
     def __init__(self, inner: LanceQuery):
         """
         Construct an AsyncQuery
@@ -2588,7 +2619,7 @@ class AsyncQuery(AsyncQueryBase):
         return AsyncFTSQuery(self._inner.nearest_to_text({"query": query}))
 
 
-class AsyncFTSQuery(AsyncQueryBase):
+class AsyncFTSQuery(AsyncStandardQuery):
     """A query for full text search for LanceDB."""
 
     def __init__(self, inner: LanceFTSQuery):
@@ -2867,7 +2898,7 @@ class AsyncVectorQueryBase:
         return self
 
 
-class AsyncVectorQuery(AsyncQueryBase, AsyncVectorQueryBase):
+class AsyncVectorQuery(AsyncStandardQuery, AsyncVectorQueryBase):
     def __init__(self, inner: LanceVectorQuery):
         """
         Construct an AsyncVectorQuery
@@ -2950,7 +2981,7 @@ class AsyncVectorQuery(AsyncQueryBase, AsyncVectorQueryBase):
         return AsyncRecordBatchReader(results, max_batch_length=max_batch_length)
 
 
-class AsyncHybridQuery(AsyncQueryBase, AsyncVectorQueryBase):
+class AsyncHybridQuery(AsyncStandardQuery, AsyncVectorQueryBase):
     """
     A query builder that performs hybrid vector and full text search.
     Results are combined and reranked based on the specified reranker.
@@ -3054,7 +3085,7 @@ class AsyncHybridQuery(AsyncQueryBase, AsyncVectorQueryBase):
             CoalesceBatchesExec: target_batch_size=1024
               GlobalLimitExec: skip=0, fetch=10
                 FilterExec: _distance@2 IS NOT NULL
-                  SortExec: TopK(fetch=10), expr=[_distance@2 ASC NULLS LAST], preserve_partitioning=[false]
+                  SortExec: TopK(fetch=10), expr=[_distance@2 ASC NULLS LAST, _rowid@1 ASC NULLS LAST], preserve_partitioning=[false]
                     KNNVectorDistance: metric=l2
                       LanceRead: uri=..., projection=[vector], ...
         <BLANKLINE>
@@ -3102,3 +3133,252 @@ class AsyncHybridQuery(AsyncQueryBase, AsyncVectorQueryBase):
         results.append(await self._inner.to_fts_query().analyze_plan())
 
         return "\n".join(results)
+
+
+class AsyncTakeQuery(AsyncQueryBase):
+    """
+    Builder for parameterizing and executing take queries.
+    """
+
+    def __init__(self, inner: LanceTakeQuery):
+        super().__init__(inner)
+
+
+class BaseQueryBuilder(object):
+    """
+    Wraps AsyncQueryBase and provides a synchronous interface
+    """
+
+    def __init__(self, inner: AsyncQueryBase):
+        self._inner = inner
+
+    def to_query_object(self) -> Query:
+        return self._inner.to_query_object()
+
+    def select(self, columns: Union[List[str], dict[str, str]]) -> Self:
+        """
+        Return only the specified columns.
+
+        By default a query will return all columns from the table.  However, this can
+        have a very significant impact on latency.  LanceDb stores data in a columnar
+        fashion.  This
+        means we can finely tune our I/O to select exactly the columns we need.
+
+        As a best practice you should always limit queries to the columns that you need.
+        If you pass in a list of column names then only those columns will be
+        returned.
+
+        You can also use this method to create new "dynamic" columns based on your
+        existing columns. For example, you may not care about "a" or "b" but instead
+        simply want "a + b".  This is often seen in the SELECT clause of an SQL query
+        (e.g. `SELECT a+b FROM my_table`).
+
+        To create dynamic columns you can pass in a dict[str, str].  A column will be
+        returned for each entry in the map.  The key provides the name of the column.
+        The value is an SQL string used to specify how the column is calculated.
+
+        For example, an SQL query might state `SELECT a + b AS combined, c`.  The
+        equivalent input to this method would be `{"combined": "a + b", "c": "c"}`.
+
+        Columns will always be returned in the order given, even if that order is
+        different than the order used when adding the data.
+        """
+        self._inner.select(columns)
+        return self
+
+    def with_row_id(self) -> Self:
+        """
+        Include the _rowid column in the results.
+        """
+        self._inner.with_row_id()
+        return self
+
+    def to_batches(
+        self,
+        *,
+        max_batch_length: Optional[int] = None,
+        timeout: Optional[timedelta] = None,
+    ) -> pa.RecordBatchReader:
+        """
+        Execute the query and return the results as an Apache Arrow RecordBatchReader.
+
+        Parameters
+        ----------
+
+        max_batch_length: Optional[int]
+            The maximum number of selected records in a single RecordBatch object.
+            If not specified, a default batch length is used.
+            It is possible for batches to be smaller than the provided length if the
+            underlying data is stored in smaller chunks.
+        timeout: Optional[timedelta]
+            The maximum time to wait for the query to complete.
+            If not specified, no timeout is applied. If the query does not
+            complete within the specified time, an error will be raised.
+        """
+        async_iter = LOOP.run(self._inner.execute(max_batch_length, timeout))
+
+        def iter_sync():
+            try:
+                while True:
+                    yield LOOP.run(async_iter.__anext__())
+            except StopAsyncIteration:
+                return
+
+        return pa.RecordBatchReader.from_batches(async_iter.schema, iter_sync())
+
+    def to_arrow(self, timeout: Optional[timedelta] = None) -> pa.Table:
+        """
+        Execute the query and collect the results into an Apache Arrow Table.
+
+        This method will collect all results into memory before returning.  If
+        you expect a large number of results, you may want to use
+        [to_batches][lancedb.query.AsyncQueryBase.to_batches]
+
+        Parameters
+        ----------
+        timeout: Optional[timedelta]
+            The maximum time to wait for the query to complete.
+            If not specified, no timeout is applied. If the query does not
+            complete within the specified time, an error will be raised.
+        """
+        return LOOP.run(self._inner.to_arrow(timeout))
+
+    def to_list(self, timeout: Optional[timedelta] = None) -> List[dict]:
+        """
+        Execute the query and return the results as a list of dictionaries.
+
+        Each list entry is a dictionary with the selected column names as keys,
+        or all table columns if `select` is not called. The vector and the "_distance"
+        fields are returned whether or not they're explicitly selected.
+
+        Parameters
+        ----------
+        timeout: Optional[timedelta]
+            The maximum time to wait for the query to complete.
+            If not specified, no timeout is applied. If the query does not
+            complete within the specified time, an error will be raised.
+        """
+        return LOOP.run(self._inner.to_list(timeout))
+
+    def to_pandas(
+        self,
+        flatten: Optional[Union[int, bool]] = None,
+        timeout: Optional[timedelta] = None,
+    ) -> "pd.DataFrame":
+        """
+        Execute the query and collect the results into a pandas DataFrame.
+
+        This method will collect all results into memory before returning.  If you
+        expect a large number of results, you may want to use
+        [to_batches][lancedb.query.AsyncQueryBase.to_batches] and convert each batch to
+        pandas separately.
+
+        Examples
+        --------
+
+        >>> import asyncio
+        >>> from lancedb import connect_async
+        >>> async def doctest_example():
+        ...     conn = await connect_async("./.lancedb")
+        ...     table = await conn.create_table("my_table", data=[{"a": 1, "b": 2}])
+        ...     async for batch in await table.query().to_batches():
+        ...         batch_df = batch.to_pandas()
+        >>> asyncio.run(doctest_example())
+
+        Parameters
+        ----------
+        flatten: Optional[Union[int, bool]]
+            If flatten is True, flatten all nested columns.
+            If flatten is an integer, flatten the nested columns up to the
+            specified depth.
+            If unspecified, do not flatten the nested columns.
+        timeout: Optional[timedelta]
+            The maximum time to wait for the query to complete.
+            If not specified, no timeout is applied. If the query does not
+            complete within the specified time, an error will be raised.
+        """
+        return LOOP.run(self._inner.to_pandas(flatten, timeout))
+
+    def to_polars(
+        self,
+        timeout: Optional[timedelta] = None,
+    ) -> "pl.DataFrame":
+        """
+        Execute the query and collect the results into a Polars DataFrame.
+
+        This method will collect all results into memory before returning.  If you
+        expect a large number of results, you may want to use
+        [to_batches][lancedb.query.AsyncQueryBase.to_batches] and convert each batch to
+        polars separately.
+
+        Parameters
+        ----------
+        timeout: Optional[timedelta]
+            The maximum time to wait for the query to complete.
+            If not specified, no timeout is applied. If the query does not
+            complete within the specified time, an error will be raised.
+
+        Examples
+        --------
+
+        >>> import asyncio
+        >>> import polars as pl
+        >>> from lancedb import connect_async
+        >>> async def doctest_example():
+        ...     conn = await connect_async("./.lancedb")
+        ...     table = await conn.create_table("my_table", data=[{"a": 1, "b": 2}])
+        ...     async for batch in await table.query().to_batches():
+        ...         batch_df = pl.from_arrow(batch)
+        >>> asyncio.run(doctest_example())
+        """
+        return LOOP.run(self._inner.to_polars(timeout))
+
+    def explain_plan(self, verbose: Optional[bool] = False):
+        """Return the execution plan for this query.
+
+        Examples
+        --------
+        >>> import asyncio
+        >>> from lancedb import connect_async
+        >>> async def doctest_example():
+        ...     conn = await connect_async("./.lancedb")
+        ...     table = await conn.create_table("my_table", [{"vector": [99, 99]}])
+        ...     query = [100, 100]
+        ...     plan = await table.query().nearest_to([1, 2]).explain_plan(True)
+        ...     print(plan)
+        >>> asyncio.run(doctest_example()) # doctest: +ELLIPSIS, +NORMALIZE_WHITESPACE
+        ProjectionExec: expr=[vector@0 as vector, _distance@2 as _distance]
+          GlobalLimitExec: skip=0, fetch=10
+            FilterExec: _distance@2 IS NOT NULL
+              SortExec: TopK(fetch=10), expr=[_distance@2 ASC NULLS LAST, _rowid@1 ASC NULLS LAST], preserve_partitioning=[false]
+                KNNVectorDistance: metric=l2
+                  LanceRead: uri=..., projection=[vector], ...
+
+        Parameters
+        ----------
+        verbose : bool, default False
+            Use a verbose output format.
+
+        Returns
+        -------
+        plan : str
+        """  # noqa: E501
+        return LOOP.run(self._inner.explain_plan(verbose))
+
+    def analyze_plan(self):
+        """Execute the query and display with runtime metrics.
+
+        Returns
+        -------
+        plan : str
+        """
+        return LOOP.run(self._inner.analyze_plan())
+
+
+class LanceTakeQueryBuilder(BaseQueryBuilder):
+    """
+    Builder for parameterizing and executing take queries.
+    """
+
+    def __init__(self, inner: AsyncTakeQuery):
+        super().__init__(inner)

python/python/lancedb/remote/__init__.py

@@ -8,7 +8,15 @@ from typing import List, Optional
 
 from lancedb import __version__
 
-__all__ = ["TimeoutConfig", "RetryConfig", "ClientConfig"]
+from .header import HeaderProvider
+
+__all__ = [
+    "TimeoutConfig",
+    "RetryConfig",
+    "TlsConfig",
+    "ClientConfig",
+    "HeaderProvider",
+]
 
 
 @dataclass
@@ -112,15 +120,43 @@ class RetryConfig:
     statuses: Optional[List[int]] = None
 
 
+@dataclass
+class TlsConfig:
+    """TLS/mTLS configuration for the remote HTTP client.
+
+    Attributes
+    ----------
+    cert_file: Optional[str]
+        Path to the client certificate file (PEM format) for mTLS authentication.
+    key_file: Optional[str]
+        Path to the client private key file (PEM format) for mTLS authentication.
+    ssl_ca_cert: Optional[str]
+        Path to the CA certificate file (PEM format) for server verification.
+    assert_hostname: bool
+        Whether to verify the hostname in the server's certificate. Default is True.
+        Set to False to disable hostname verification (use with caution).
+    """
+
+    cert_file: Optional[str] = None
+    key_file: Optional[str] = None
+    ssl_ca_cert: Optional[str] = None
+    assert_hostname: bool = True
+
+
 @dataclass
 class ClientConfig:
     user_agent: str = f"LanceDB-Python-Client/{__version__}"
     retry_config: RetryConfig = field(default_factory=RetryConfig)
     timeout_config: Optional[TimeoutConfig] = field(default_factory=TimeoutConfig)
     extra_headers: Optional[dict] = None
+    id_delimiter: Optional[str] = None
+    tls_config: Optional[TlsConfig] = None
+    header_provider: Optional["HeaderProvider"] = None
 
     def __post_init__(self):
         if isinstance(self.retry_config, dict):
             self.retry_config = RetryConfig(**self.retry_config)
         if isinstance(self.timeout_config, dict):
             self.timeout_config = TimeoutConfig(**self.timeout_config)
+        if isinstance(self.tls_config, dict):
+            self.tls_config = TlsConfig(**self.tls_config)

python/python/lancedb/remote/db.py

@@ -96,14 +96,73 @@ class RemoteDBConnection(DBConnection):
     def __repr__(self) -> str:
         return f"RemoteConnect(name={self.db_name})"
 
+    @override
+    def list_namespaces(
+        self,
+        namespace: List[str] = [],
+        page_token: Optional[str] = None,
+        limit: int = 10,
+    ) -> Iterable[str]:
+        """List immediate child namespace names in the given namespace.
+
+        Parameters
+        ----------
+        namespace: List[str], optional
+            The parent namespace to list namespaces in.
+            None or empty list represents root namespace.
+        page_token: str, optional
+            The token to use for pagination. If not present, start from the beginning.
+        limit: int, default 10
+            The size of the page to return.
+
+        Returns
+        -------
+        Iterable of str
+            List of immediate child namespace names
+        """
+        return LOOP.run(
+            self._conn.list_namespaces(
+                namespace=namespace, page_token=page_token, limit=limit
+            )
+        )
+
+    @override
+    def create_namespace(self, namespace: List[str]) -> None:
+        """Create a new namespace.
+
+        Parameters
+        ----------
+        namespace: List[str]
+            The namespace identifier to create.
+        """
+        LOOP.run(self._conn.create_namespace(namespace=namespace))
+
+    @override
+    def drop_namespace(self, namespace: List[str]) -> None:
+        """Drop a namespace.
+
+        Parameters
+        ----------
+        namespace: List[str]
+            The namespace identifier to drop.
+        """
+        return LOOP.run(self._conn.drop_namespace(namespace=namespace))
+
     @override
     def table_names(
-        self, page_token: Optional[str] = None, limit: int = 10
+        self,
+        page_token: Optional[str] = None,
+        limit: int = 10,
+        *,
+        namespace: List[str] = [],
     ) -> Iterable[str]:
         """List the names of all tables in the database.
 
         Parameters
         ----------
+        namespace: List[str], default []
+            The namespace to list tables in.
+            Empty list represents root namespace.
         page_token: str
             The last token to start the new page.
         limit: int, default 10
@@ -113,13 +172,18 @@ class RemoteDBConnection(DBConnection):
         -------
         An iterator of table names.
         """
-        return LOOP.run(self._conn.table_names(start_after=page_token, limit=limit))
+        return LOOP.run(
+            self._conn.table_names(
+                namespace=namespace, start_after=page_token, limit=limit
+            )
+        )
 
     @override
     def open_table(
         self,
         name: str,
         *,
+        namespace: List[str] = [],
         storage_options: Optional[Dict[str, str]] = None,
         index_cache_size: Optional[int] = None,
     ) -> Table:
@@ -129,6 +193,9 @@ class RemoteDBConnection(DBConnection):
         ----------
         name: str
             The name of the table.
+        namespace: List[str], optional
+            The namespace to open the table from.
+            None or empty list represents root namespace.
 
         Returns
         -------
@@ -142,7 +209,54 @@ class RemoteDBConnection(DBConnection):
                 " (there is no local cache to configure)"
             )
 
-        table = LOOP.run(self._conn.open_table(name))
+        table = LOOP.run(self._conn.open_table(name, namespace=namespace))
+        return RemoteTable(table, self.db_name)
+
+    def clone_table(
+        self,
+        target_table_name: str,
+        source_uri: str,
+        *,
+        target_namespace: List[str] = [],
+        source_version: Optional[int] = None,
+        source_tag: Optional[str] = None,
+        is_shallow: bool = True,
+    ) -> Table:
+        """Clone a table from a source table.
+
+        Parameters
+        ----------
+        target_table_name: str
+            The name of the target table to create.
+        source_uri: str
+            The URI of the source table to clone from.
+        target_namespace: List[str], optional
+            The namespace for the target table.
+            None or empty list represents root namespace.
+        source_version: int, optional
+            The version of the source table to clone.
+        source_tag: str, optional
+            The tag of the source table to clone.
+        is_shallow: bool, default True
+            Whether to perform a shallow clone (True) or deep clone (False).
+            Currently only shallow clone is supported.
+
+        Returns
+        -------
+        A RemoteTable object representing the cloned table.
+        """
+        from .table import RemoteTable
+
+        table = LOOP.run(
+            self._conn.clone_table(
+                target_table_name,
+                source_uri,
+                target_namespace=target_namespace,
+                source_version=source_version,
+                source_tag=source_tag,
+                is_shallow=is_shallow,
+            )
+        )
         return RemoteTable(table, self.db_name)
 
     @override
@@ -155,6 +269,8 @@ class RemoteDBConnection(DBConnection):
         fill_value: float = 0.0,
         mode: Optional[str] = None,
         embedding_functions: Optional[List[EmbeddingFunctionConfig]] = None,
+        *,
+        namespace: List[str] = [],
     ) -> Table:
         """Create a [Table][lancedb.table.Table] in the database.
 
@@ -162,6 +278,9 @@ class RemoteDBConnection(DBConnection):
         ----------
         name: str
             The name of the table.
+        namespace: List[str], optional
+            The namespace to create the table in.
+            None or empty list represents root namespace.
         data: The data to initialize the table, *optional*
             User must provide at least one of `data` or `schema`.
             Acceptable types are:
@@ -262,6 +381,7 @@ class RemoteDBConnection(DBConnection):
             self._conn.create_table(
                 name,
                 data,
+                namespace=namespace,
                 mode=mode,
                 schema=schema,
                 on_bad_vectors=on_bad_vectors,
@@ -271,18 +391,27 @@ class RemoteDBConnection(DBConnection):
         return RemoteTable(table, self.db_name)
 
     @override
-    def drop_table(self, name: str):
+    def drop_table(self, name: str, namespace: List[str] = []):
         """Drop a table from the database.
 
         Parameters
         ----------
         name: str
             The name of the table.
+        namespace: List[str], optional
+            The namespace to drop the table from.
+            None or empty list represents root namespace.
         """
-        LOOP.run(self._conn.drop_table(name))
+        LOOP.run(self._conn.drop_table(name, namespace=namespace))
 
     @override
-    def rename_table(self, cur_name: str, new_name: str):
+    def rename_table(
+        self,
+        cur_name: str,
+        new_name: str,
+        cur_namespace: List[str] = [],
+        new_namespace: List[str] = [],
+    ):
         """Rename a table in the database.
 
         Parameters
@@ -292,7 +421,14 @@ class RemoteDBConnection(DBConnection):
         new_name: str
             The new name of the table.
         """
-        LOOP.run(self._conn.rename_table(cur_name, new_name))
+        LOOP.run(
+            self._conn.rename_table(
+                cur_name,
+                new_name,
+                cur_namespace=cur_namespace,
+                new_namespace=new_namespace,
+            )
+        )
 
     async def close(self):
         """Close the connection to the database."""

python/python/lancedb/remote/header.py

@@ -0,0 +1,180 @@
+# SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright The LanceDB Authors
+
+"""Header providers for LanceDB remote connections.
+
+This module provides a flexible header management framework for LanceDB remote
+connections, allowing users to implement custom header strategies for
+authentication, request tracking, custom metadata, or any other header-based
+requirements.
+
+The module includes the HeaderProvider abstract base class and example implementations
+(StaticHeaderProvider and OAuthProvider) that demonstrate common patterns.
+
+The HeaderProvider interface is designed to be called before each request to the remote
+server, enabling dynamic header scenarios where values may need to be
+refreshed, rotated, or computed on-demand.
+"""
+
+from abc import ABC, abstractmethod
+from typing import Dict, Optional, Callable, Any
+import time
+import threading
+
+
+class HeaderProvider(ABC):
+    """Abstract base class for providing custom headers for each request.
+
+    Users can implement this interface to provide dynamic headers for various purposes
+    such as authentication (OAuth tokens, API keys), request tracking (correlation IDs),
+    custom metadata, or any other header-based requirements. The provider is called
+    before each request to ensure fresh header values are always used.
+
+    Error Handling
+    --------------
+    If get_headers() raises an exception, the request will fail. Implementations
+    should handle recoverable errors internally (e.g., retry token refresh) and
+    only raise exceptions for unrecoverable errors.
+    """
+
+    @abstractmethod
+    def get_headers(self) -> Dict[str, str]:
+        """Get the latest headers to be added to requests.
+
+        This method is called before each request to the remote LanceDB server.
+        Implementations should return headers that will be merged with existing headers.
+
+        Returns
+        -------
+        Dict[str, str]
+            Dictionary of header names to values to add to the request.
+
+        Raises
+        ------
+        Exception
+            If unable to fetch headers, the exception will be propagated
+            and the request will fail.
+        """
+        pass
+
+
+class StaticHeaderProvider(HeaderProvider):
+    """Example implementation: A simple header provider that returns static headers.
+
+    This is an example implementation showing how to create a HeaderProvider
+    for cases where headers don't change during the session. Users can use this
+    as a reference for implementing their own providers.
+
+    Parameters
+    ----------
+    headers : Dict[str, str]
+        Static headers to return for every request.
+    """
+
+    def __init__(self, headers: Dict[str, str]):
+        """Initialize with static headers.
+
+        Parameters
+        ----------
+        headers : Dict[str, str]
+            Headers to return for every request.
+        """
+        self._headers = headers.copy()
+
+    def get_headers(self) -> Dict[str, str]:
+        """Return the static headers.
+
+        Returns
+        -------
+        Dict[str, str]
+            Copy of the static headers.
+        """
+        return self._headers.copy()
+
+
+class OAuthProvider(HeaderProvider):
+    """Example implementation: OAuth token provider with automatic refresh.
+
+    This is an example implementation showing how to manage OAuth tokens
+    with automatic refresh when they expire. Users can use this as a reference
+    for implementing their own OAuth or token-based authentication providers.
+
+    Parameters
+    ----------
+    token_fetcher : Callable[[], Dict[str, Any]]
+        Function that fetches a new token. Should return a dict with
+        'access_token' and optionally 'expires_in' (seconds until expiration).
+    refresh_buffer_seconds : int, optional
+        Number of seconds before expiration to trigger refresh. Default is 300
+        (5 minutes).
+    """
+
+    def __init__(
+        self, token_fetcher: Callable[[], Any], refresh_buffer_seconds: int = 300
+    ):
+        """Initialize the OAuth provider.
+
+        Parameters
+        ----------
+        token_fetcher : Callable[[], Any]
+            Function to fetch new tokens. Should return dict with
+            'access_token' and optionally 'expires_in'.
+        refresh_buffer_seconds : int, optional
+            Seconds before expiry to refresh token. Default 300.
+        """
+        self._token_fetcher = token_fetcher
+        self._refresh_buffer = refresh_buffer_seconds
+        self._current_token: Optional[str] = None
+        self._token_expires_at: Optional[float] = None
+        self._refresh_lock = threading.Lock()
+
+    def _refresh_token_if_needed(self) -> None:
+        """Refresh the token if it's expired or close to expiring."""
+        with self._refresh_lock:
+            # Check again inside the lock in case another thread refreshed
+            if self._needs_refresh():
+                token_data = self._token_fetcher()
+
+                self._current_token = token_data.get("access_token")
+                if not self._current_token:
+                    raise ValueError("Token fetcher did not return 'access_token'")
+
+                # Set expiration if provided
+                expires_in = token_data.get("expires_in")
+                if expires_in:
+                    self._token_expires_at = time.time() + expires_in
+                else:
+                    # Token doesn't expire or expiration unknown
+                    self._token_expires_at = None
+
+    def _needs_refresh(self) -> bool:
+        """Check if token needs refresh."""
+        if self._current_token is None:
+            return True
+
+        if self._token_expires_at is None:
+            # No expiration info, assume token is valid
+            return False
+
+        # Refresh if we're within the buffer time of expiration
+        return time.time() >= (self._token_expires_at - self._refresh_buffer)
+
+    def get_headers(self) -> Dict[str, str]:
+        """Get OAuth headers, refreshing token if needed.
+
+        Returns
+        -------
+        Dict[str, str]
+            Headers with Bearer token authorization.
+
+        Raises
+        ------
+        Exception
+            If unable to fetch or refresh token.
+        """
+        self._refresh_token_if_needed()
+
+        if not self._current_token:
+            raise RuntimeError("Failed to obtain OAuth token")
+
+        return {"Authorization": f"Bearer {self._current_token}"}

python/python/lancedb/remote/table.py

@@ -26,7 +26,7 @@ from lancedb.common import DATA, VEC, VECTOR_COLUMN_NAME
 from lancedb.merge import LanceMergeInsertBuilder
 from lancedb.embeddings import EmbeddingFunctionRegistry
 
-from ..query import LanceVectorQueryBuilder, LanceQueryBuilder
+from ..query import LanceVectorQueryBuilder, LanceQueryBuilder, LanceTakeQueryBuilder
 from ..table import AsyncTable, IndexStatistics, Query, Table, Tags
 
 
@@ -115,6 +115,7 @@ class RemoteTable(Table):
         *,
         replace: bool = False,
         wait_timeout: timedelta = None,
+        name: Optional[str] = None,
     ):
         """Creates a scalar index
         Parameters
@@ -139,7 +140,11 @@ class RemoteTable(Table):
 
         LOOP.run(
             self._table.create_index(
-                column, config=config, replace=replace, wait_timeout=wait_timeout
+                column,
+                config=config,
+                replace=replace,
+                wait_timeout=wait_timeout,
+                name=name,
             )
         )
 
@@ -161,6 +166,7 @@ class RemoteTable(Table):
         ngram_min_length: int = 3,
         ngram_max_length: int = 3,
         prefix_only: bool = False,
+        name: Optional[str] = None,
     ):
         config = FTS(
             with_position=with_position,
@@ -177,7 +183,11 @@ class RemoteTable(Table):
         )
         LOOP.run(
             self._table.create_index(
-                column, config=config, replace=replace, wait_timeout=wait_timeout
+                column,
+                config=config,
+                replace=replace,
+                wait_timeout=wait_timeout,
+                name=name,
             )
         )
 
@@ -194,6 +204,8 @@ class RemoteTable(Table):
         wait_timeout: Optional[timedelta] = None,
         *,
         num_bits: int = 8,
+        name: Optional[str] = None,
+        train: bool = True,
     ):
         """Create an index on the table.
         Currently, the only parameters that matter are
@@ -270,7 +282,11 @@ class RemoteTable(Table):
 
         LOOP.run(
             self._table.create_index(
-                vector_column_name, config=config, wait_timeout=wait_timeout
+                vector_column_name,
+                config=config,
+                wait_timeout=wait_timeout,
+                name=name,
+                train=train,
             )
         )
 
@@ -617,6 +633,12 @@ class RemoteTable(Table):
     def stats(self):
         return LOOP.run(self._table.stats())
 
+    def take_offsets(self, offsets: list[int]) -> LanceTakeQueryBuilder:
+        return LanceTakeQueryBuilder(self._table.take_offsets(offsets))
+
+    def take_row_ids(self, row_ids: list[int]) -> LanceTakeQueryBuilder:
+        return LanceTakeQueryBuilder(self._table.take_row_ids(row_ids))
+
     def uses_v2_manifest_paths(self) -> bool:
         raise NotImplementedError(
             "uses_v2_manifest_paths() is not supported on the LanceDB Cloud"

python/python/lancedb/rerankers/__init__.py

@@ -9,6 +9,7 @@ from .linear_combination import LinearCombinationReranker
 from .openai import OpenaiReranker
 from .jinaai import JinaReranker
 from .rrf import RRFReranker
+from .mrr import MRRReranker
 from .answerdotai import AnswerdotaiRerankers
 from .voyageai import VoyageAIReranker
 
@@ -23,4 +24,5 @@ __all__ = [
     "RRFReranker",
     "AnswerdotaiRerankers",
     "VoyageAIReranker",
+    "MRRReranker",
 ]

python/python/lancedb/rerankers/mrr.py

@@ -0,0 +1,169 @@
+# SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright The LanceDB Authors
+
+
+from typing import Union, List, TYPE_CHECKING
+import pyarrow as pa
+import numpy as np
+
+from collections import defaultdict
+from .base import Reranker
+
+if TYPE_CHECKING:
+    from ..table import LanceVectorQueryBuilder
+
+
+class MRRReranker(Reranker):
+    """
+    Reranks the results using Mean Reciprocal Rank (MRR) algorithm based
+    on the scores of vector and FTS search.
+    Algorithm reference - https://en.wikipedia.org/wiki/Mean_reciprocal_rank
+
+    MRR calculates the average of reciprocal ranks across different search results.
+    For each document, it computes the reciprocal of its rank in each system,
+    then takes the mean of these reciprocal ranks as the final score.
+
+    Parameters
+    ----------
+    weight_vector : float, default 0.5
+        Weight for vector search results (0.0 to 1.0)
+    weight_fts : float, default 0.5
+        Weight for FTS search results (0.0 to 1.0)
+        Note: weight_vector + weight_fts should equal 1.0
+    return_score : str, default "relevance"
+        Options are "relevance" or "all"
+        The type of score to return. If "relevance", will return only the relevance
+        score. If "all", will return all scores from the vector and FTS search along
+        with the relevance score.
+    """
+
+    def __init__(
+        self,
+        weight_vector: float = 0.5,
+        weight_fts: float = 0.5,
+        return_score="relevance",
+    ):
+        if not (0.0 <= weight_vector <= 1.0):
+            raise ValueError("weight_vector must be between 0.0 and 1.0")
+        if not (0.0 <= weight_fts <= 1.0):
+            raise ValueError("weight_fts must be between 0.0 and 1.0")
+        if abs(weight_vector + weight_fts - 1.0) > 1e-6:
+            raise ValueError("weight_vector + weight_fts must equal 1.0")
+
+        super().__init__(return_score)
+        self.weight_vector = weight_vector
+        self.weight_fts = weight_fts
+
+    def rerank_hybrid(
+        self,
+        query: str,  # noqa: F821
+        vector_results: pa.Table,
+        fts_results: pa.Table,
+    ):
+        vector_ids = vector_results["_rowid"].to_pylist() if vector_results else []
+        fts_ids = fts_results["_rowid"].to_pylist() if fts_results else []
+
+        # Maps result_id to list of (type, reciprocal_rank)
+        mrr_score_map = defaultdict(list)
+
+        if vector_ids:
+            for rank, result_id in enumerate(vector_ids, 1):
+                reciprocal_rank = 1.0 / rank
+                mrr_score_map[result_id].append(("vector", reciprocal_rank))
+
+        if fts_ids:
+            for rank, result_id in enumerate(fts_ids, 1):
+                reciprocal_rank = 1.0 / rank
+                mrr_score_map[result_id].append(("fts", reciprocal_rank))
+
+        final_mrr_scores = {}
+        for result_id, scores in mrr_score_map.items():
+            vector_rr = 0.0
+            fts_rr = 0.0
+
+            for score_type, reciprocal_rank in scores:
+                if score_type == "vector":
+                    vector_rr = reciprocal_rank
+                elif score_type == "fts":
+                    fts_rr = reciprocal_rank
+
+            # If a document doesn't appear, its reciprocal rank is 0
+            weighted_mrr = self.weight_vector * vector_rr + self.weight_fts * fts_rr
+            final_mrr_scores[result_id] = weighted_mrr
+
+        combined_results = self.merge_results(vector_results, fts_results)
+        combined_row_ids = combined_results["_rowid"].to_pylist()
+        relevance_scores = [final_mrr_scores[row_id] for row_id in combined_row_ids]
+        combined_results = combined_results.append_column(
+            "_relevance_score", pa.array(relevance_scores, type=pa.float32())
+        )
+        combined_results = combined_results.sort_by(
+            [("_relevance_score", "descending")]
+        )
+
+        if self.score == "relevance":
+            combined_results = self._keep_relevance_score(combined_results)
+
+        return combined_results
+
+    def rerank_multivector(
+        self,
+        vector_results: Union[List[pa.Table], List["LanceVectorQueryBuilder"]],
+        query: str = None,
+        deduplicate: bool = True,  # noqa: F821
+    ):
+        """
+        Reranks the results from multiple vector searches using MRR algorithm.
+        Each vector search result is treated as a separate ranking system,
+        and MRR calculates the mean of reciprocal ranks across all systems.
+        This cannot reuse rerank_hybrid because MRR semantics require treating
+        each vector result as a separate ranking system.
+        """
+        if not all(isinstance(v, type(vector_results[0])) for v in vector_results):
+            raise ValueError(
+                "All elements in vector_results should be of the same type"
+            )
+
+        # avoid circular import
+        if type(vector_results[0]).__name__ == "LanceVectorQueryBuilder":
+            vector_results = [result.to_arrow() for result in vector_results]
+        elif not isinstance(vector_results[0], pa.Table):
+            raise ValueError(
+                "vector_results should be a list of pa.Table or LanceVectorQueryBuilder"
+            )
+
+        if not all("_rowid" in result.column_names for result in vector_results):
+            raise ValueError(
+                "'_rowid' is required for deduplication. \
+                    add _rowid to search results like this: \
+                    `search().with_row_id(True)`"
+            )
+
+        mrr_score_map = defaultdict(list)
+
+        for result_table in vector_results:
+            result_ids = result_table["_rowid"].to_pylist()
+            for rank, result_id in enumerate(result_ids, 1):
+                reciprocal_rank = 1.0 / rank
+                mrr_score_map[result_id].append(reciprocal_rank)
+
+        final_mrr_scores = {}
+        for result_id, reciprocal_ranks in mrr_score_map.items():
+            mean_rr = np.mean(reciprocal_ranks)
+            final_mrr_scores[result_id] = mean_rr
+
+        combined = pa.concat_tables(vector_results, **self._concat_tables_args)
+        combined = self._deduplicate(combined)
+
+        combined_row_ids = combined["_rowid"].to_pylist()
+
+        relevance_scores = [final_mrr_scores[row_id] for row_id in combined_row_ids]
+        combined = combined.append_column(
+            "_relevance_score", pa.array(relevance_scores, type=pa.float32())
+        )
+        combined = combined.sort_by([("_relevance_score", "descending")])
+
+        if self.score == "relevance":
+            combined = self._keep_relevance_score(combined)
+
+        return combined

python/python/lancedb/table.py

@@ -51,6 +51,7 @@ from .query import (
     AsyncFTSQuery,
     AsyncHybridQuery,
     AsyncQuery,
+    AsyncTakeQuery,
     AsyncVectorQuery,
     FullTextQuery,
     LanceEmptyQueryBuilder,
@@ -58,6 +59,7 @@ from .query import (
     LanceHybridQueryBuilder,
     LanceQueryBuilder,
     LanceVectorQueryBuilder,
+    LanceTakeQueryBuilder,
     Query,
 )
 from .util import (
@@ -687,6 +689,9 @@ class Table(ABC):
         sample_rate: int = 256,
         m: int = 20,
         ef_construction: int = 300,
+        name: Optional[str] = None,
+        train: bool = True,
+        target_partition_size: Optional[int] = None,
     ):
         """Create an index on the table.
 
@@ -719,6 +724,11 @@ class Table(ABC):
             Only 4 and 8 are supported.
         wait_timeout: timedelta, optional
             The timeout to wait if indexing is asynchronous.
+        name: str, optional
+            The name of the index. If not provided, a default name will be generated.
+        train: bool, default True
+            Whether to train the index with existing data. Vector indices always train
+            with existing data.
         """
         raise NotImplementedError
 
@@ -774,6 +784,7 @@ class Table(ABC):
         replace: bool = True,
         index_type: ScalarIndexType = "BTREE",
         wait_timeout: Optional[timedelta] = None,
+        name: Optional[str] = None,
     ):
         """Create a scalar index on a column.
 
@@ -788,6 +799,8 @@ class Table(ABC):
             The type of index to create.
         wait_timeout: timedelta, optional
             The timeout to wait if indexing is asynchronous.
+        name: str, optional
+            The name of the index. If not provided, a default name will be generated.
         Examples
         --------
 
@@ -850,6 +863,7 @@ class Table(ABC):
         ngram_max_length: int = 3,
         prefix_only: bool = False,
         wait_timeout: Optional[timedelta] = None,
+        name: Optional[str] = None,
     ):
         """Create a full-text search index on the table.
 
@@ -914,6 +928,8 @@ class Table(ABC):
             Whether to only index the prefix of the token for ngram tokenizer.
         wait_timeout: timedelta, optional
             The timeout to wait if indexing is asynchronous.
+        name: str, optional
+            The name of the index. If not provided, a default name will be generated.
         """
         raise NotImplementedError
 
@@ -1103,6 +1119,120 @@ class Table(ABC):
         """
         raise NotImplementedError
 
+    @abstractmethod
+    def take_offsets(
+        self, offsets: list[int], *, with_row_id: bool = False
+    ) -> LanceTakeQueryBuilder:
+        """
+        Take a list of offsets from the table.
+
+        Offsets are 0-indexed and relative to the current version of the table.  Offsets
+        are not stable.  A row with an offset of N may have a different offset in a
+        different version of the table (e.g. if an earlier row is deleted).
+
+        Offsets are mostly useful for sampling as the set of all valid offsets is easily
+        known in advance to be [0, len(table)).
+
+        No guarantees are made regarding the order in which results are returned.  If
+        you desire an output order that matches the order of the given offsets, you will
+        need to add the row offset column to the output and align it yourself.
+
+        Parameters
+        ----------
+        offsets: list[int]
+            The offsets to take.
+
+        Returns
+        -------
+        pa.RecordBatch
+            A record batch containing the rows at the given offsets.
+        """
+
+    def __getitems__(self, offsets: list[int]) -> pa.RecordBatch:
+        """
+        Take a list of offsets from the table and return as a record batch.
+
+        This method uses the `take_offsets` method to take the rows.  However, it
+        aligns the offsets to the passed in offsets.  This means the return type
+        is a record batch (and so users should take care not to pass in too many
+        offsets)
+
+        Note: this method is primarily intended to fulfill the Dataset contract
+        for pytorch.
+
+        Parameters
+        ----------
+        offsets: list[int]
+            The offsets to take.
+
+        Returns
+        -------
+        pa.RecordBatch
+            A record batch containing the rows at the given offsets.
+        """
+        # We don't know the order of the results at all.  So we calculate a permutation
+        # for ordering the given offsets.  Then we load the data with the _rowoffset
+        # column.  Then we sort by _rowoffset and apply the inverse of the permutation
+        # that we calculated.
+        #
+        # Note: this is potentially a lot of memory copy if we're operating on large
+        # batches :(
+        num_offsets = len(offsets)
+        indices = list(range(num_offsets))
+        permutation = sorted(indices, key=lambda idx: offsets[idx])
+        permutation_inv = [0] * num_offsets
+        for i in range(num_offsets):
+            permutation_inv[permutation[i]] = i
+
+        columns = self.schema.names
+        columns.append("_rowoffset")
+        tbl = (
+            self.take_offsets(offsets)
+            .select(columns)
+            .to_arrow()
+            .sort_by("_rowoffset")
+            .take(permutation_inv)
+            .combine_chunks()
+            .drop_columns(["_rowoffset"])
+        )
+
+        return tbl
+
+    @abstractmethod
+    def take_row_ids(
+        self, row_ids: list[int], *, with_row_id: bool = False
+    ) -> LanceTakeQueryBuilder:
+        """
+        Take a list of row ids from the table.
+
+        Row ids are not stable and are relative to the current version of the table.
+        They can change due to compaction and updates.
+
+        No guarantees are made regarding the order in which results are returned.  If
+        you desire an output order that matches the order of the given ids, you will
+        need to add the row id column to the output and align it yourself.
+
+        Unlike offsets, row ids are not 0-indexed and no assumptions should be made
+        about the possible range of row ids.  In order to use this method you must
+        first obtain the row ids by scanning or searching the table.
+
+        Even so, row ids are more stable than offsets and can be useful in some
+        situations.
+
+        There is an ongoing effort to make row ids stable which is tracked at
+        https://github.com/lancedb/lancedb/issues/1120
+
+        Parameters
+        ----------
+        row_ids: list[int]
+            The row ids to take.
+
+        Returns
+        -------
+        AsyncTakeQuery
+            A query object that can be executed to get the rows.
+        """
+
     @abstractmethod
     def _execute_query(
         self,
@@ -1340,10 +1470,7 @@ class Table(ABC):
             be deleted unless they are at least 7 days old. If delete_unverified is True
             then these files will be deleted regardless of their age.
         retrain: bool, default False
-            If True, retrain the vector indices, this would refine the IVF clustering
-            and quantization, which may improve the search accuracy. It's faster than
-            re-creating the index from scratch, so it's recommended to try this first,
-            when the data distribution has changed significantly.
+            This parameter is no longer used and is deprecated.
 
         Experimental API
         ----------------
@@ -1577,13 +1704,16 @@ class LanceTable(Table):
         connection: "LanceDBConnection",
         name: str,
         *,
+        namespace: List[str] = [],
         storage_options: Optional[Dict[str, str]] = None,
         index_cache_size: Optional[int] = None,
     ):
         self._conn = connection
+        self._namespace = namespace
         self._table = LOOP.run(
             connection._conn.open_table(
                 name,
+                namespace=namespace,
                 storage_options=storage_options,
                 index_cache_size=index_cache_size,
             )
@@ -1594,8 +1724,8 @@ class LanceTable(Table):
         return self._table.name
 
     @classmethod
-    def open(cls, db, name, **kwargs):
-        tbl = cls(db, name, **kwargs)
+    def open(cls, db, name, *, namespace: List[str] = [], **kwargs):
+        tbl = cls(db, name, namespace=namespace, **kwargs)
 
         # check the dataset exists
         try:
@@ -1648,6 +1778,12 @@ class LanceTable(Table):
         """Get the current version of the table"""
         return LOOP.run(self._table.version())
 
+    def take_offsets(self, offsets: list[int]) -> LanceTakeQueryBuilder:
+        return LanceTakeQueryBuilder(self._table.take_offsets(offsets))
+
+    def take_row_ids(self, row_ids: list[int]) -> LanceTakeQueryBuilder:
+        return LanceTakeQueryBuilder(self._table.take_row_ids(row_ids))
+
     @property
     def tags(self) -> Tags:
         """Tag management for the table.
@@ -1861,6 +1997,10 @@ class LanceTable(Table):
         sample_rate: int = 256,
         m: int = 20,
         ef_construction: int = 300,
+        *,
+        name: Optional[str] = None,
+        train: bool = True,
+        target_partition_size: Optional[int] = None,
     ):
         """Create an index on the table."""
         if accelerator is not None:
@@ -1877,6 +2017,7 @@ class LanceTable(Table):
                 num_bits=num_bits,
                 m=m,
                 ef_construction=ef_construction,
+                target_partition_size=target_partition_size,
             )
             self.checkout_latest()
             return
@@ -1886,6 +2027,7 @@ class LanceTable(Table):
                 num_partitions=num_partitions,
                 max_iterations=max_iterations,
                 sample_rate=sample_rate,
+                target_partition_size=target_partition_size,
             )
         elif index_type == "IVF_PQ":
             config = IvfPq(
@@ -1895,6 +2037,7 @@ class LanceTable(Table):
                 num_bits=num_bits,
                 max_iterations=max_iterations,
                 sample_rate=sample_rate,
+                target_partition_size=target_partition_size,
             )
         elif index_type == "IVF_HNSW_PQ":
             config = HnswPq(
@@ -1906,6 +2049,7 @@ class LanceTable(Table):
                 sample_rate=sample_rate,
                 m=m,
                 ef_construction=ef_construction,
+                target_partition_size=target_partition_size,
             )
         elif index_type == "IVF_HNSW_SQ":
             config = HnswSq(
@@ -1915,6 +2059,7 @@ class LanceTable(Table):
                 sample_rate=sample_rate,
                 m=m,
                 ef_construction=ef_construction,
+                target_partition_size=target_partition_size,
             )
         else:
             raise ValueError(f"Unknown index type {index_type}")
@@ -1924,6 +2069,8 @@ class LanceTable(Table):
                 vector_column_name,
                 replace=replace,
                 config=config,
+                name=name,
+                train=train,
             )
         )
 
@@ -1968,6 +2115,7 @@ class LanceTable(Table):
         *,
         replace: bool = True,
         index_type: ScalarIndexType = "BTREE",
+        name: Optional[str] = None,
     ):
         if index_type == "BTREE":
             config = BTree()
@@ -1978,7 +2126,7 @@ class LanceTable(Table):
         else:
             raise ValueError(f"Unknown index type {index_type}")
         return LOOP.run(
-            self._table.create_index(column, replace=replace, config=config)
+            self._table.create_index(column, replace=replace, config=config, name=name)
         )
 
     def create_fts_index(
@@ -2002,6 +2150,7 @@ class LanceTable(Table):
         ngram_min_length: int = 3,
         ngram_max_length: int = 3,
         prefix_only: bool = False,
+        name: Optional[str] = None,
     ):
         if not use_tantivy:
             if not isinstance(field_names, str):
@@ -2039,6 +2188,7 @@ class LanceTable(Table):
                     field_names,
                     replace=replace,
                     config=config,
+                    name=name,
                 )
             )
             return
@@ -2405,6 +2555,7 @@ class LanceTable(Table):
         fill_value: float = 0.0,
         embedding_functions: Optional[List[EmbeddingFunctionConfig]] = None,
         *,
+        namespace: List[str] = [],
         storage_options: Optional[Dict[str, str | bool]] = None,
         data_storage_version: Optional[str] = None,
         enable_v2_manifest_paths: Optional[bool] = None,
@@ -2464,6 +2615,7 @@ class LanceTable(Table):
         """
         self = cls.__new__(cls)
         self._conn = db
+        self._namespace = namespace
 
         if data_storage_version is not None:
             warnings.warn(
@@ -2496,6 +2648,7 @@ class LanceTable(Table):
                 on_bad_vectors=on_bad_vectors,
                 fill_value=fill_value,
                 embedding_functions=embedding_functions,
+                namespace=namespace,
                 storage_options=storage_options,
             )
         )
@@ -2679,10 +2832,7 @@ class LanceTable(Table):
             be deleted unless they are at least 7 days old. If delete_unverified is True
             then these files will be deleted regardless of their age.
         retrain: bool, default False
-            If True, retrain the vector indices, this would refine the IVF clustering
-            and quantization, which may improve the search accuracy. It's faster than
-            re-creating the index from scratch, so it's recommended to try this first,
-            when the data distribution has changed significantly.
+            This parameter is no longer used and is deprecated.
 
         Experimental API
         ----------------
@@ -3183,6 +3333,8 @@ class AsyncTable:
             Union[IvfFlat, IvfPq, HnswPq, HnswSq, BTree, Bitmap, LabelList, FTS]
         ] = None,
         wait_timeout: Optional[timedelta] = None,
+        name: Optional[str] = None,
+        train: bool = True,
     ):
         """Create an index to speed up queries
 
@@ -3209,6 +3361,11 @@ class AsyncTable:
             creating an index object.
         wait_timeout: timedelta, optional
             The timeout to wait if indexing is asynchronous.
+        name: str, optional
+            The name of the index. If not provided, a default name will be generated.
+        train: bool, default True
+            Whether to train the index with existing data. Vector indices always train
+            with existing data.
         """
         if config is not None:
             if not isinstance(
@@ -3220,7 +3377,12 @@ class AsyncTable:
                 )
         try:
             await self._inner.create_index(
-                column, index=config, replace=replace, wait_timeout=wait_timeout
+                column,
+                index=config,
+                replace=replace,
+                wait_timeout=wait_timeout,
+                name=name,
+                train=train,
             )
         except ValueError as e:
             if "not support the requested language" in str(e):
@@ -4030,6 +4192,58 @@ class AsyncTable:
         """
         await self._inner.restore(version)
 
+    def take_offsets(self, offsets: list[int]) -> AsyncTakeQuery:
+        """
+        Take a list of offsets from the table.
+
+        Offsets are 0-indexed and relative to the current version of the table.  Offsets
+        are not stable.  A row with an offset of N may have a different offset in a
+        different version of the table (e.g. if an earlier row is deleted).
+
+        Offsets are mostly useful for sampling as the set of all valid offsets is easily
+        known in advance to be [0, len(table)).
+
+        Parameters
+        ----------
+        offsets: list[int]
+            The offsets to take.
+
+        Returns
+        -------
+        pa.RecordBatch
+            A record batch containing the rows at the given offsets.
+        """
+        return AsyncTakeQuery(self._inner.take_offsets(offsets))
+
+    def take_row_ids(self, row_ids: list[int]) -> AsyncTakeQuery:
+        """
+        Take a list of row ids from the table.
+
+        Row ids are not stable and are relative to the current version of the table.
+        They can change due to compaction and updates.
+
+        Unlike offsets, row ids are not 0-indexed and no assumptions should be made
+        about the possible range of row ids.  In order to use this method you must
+        first obtain the row ids by scanning or searching the table.
+
+        Even so, row ids are more stable than offsets and can be useful in some
+        situations.
+
+        There is an ongoing effort to make row ids stable which is tracked at
+        https://github.com/lancedb/lancedb/issues/1120
+
+        Parameters
+        ----------
+        row_ids: list[int]
+            The row ids to take.
+
+        Returns
+        -------
+        AsyncTakeQuery
+            A query object that can be executed to get the rows.
+        """
+        return AsyncTakeQuery(self._inner.take_row_ids(row_ids))
+
     @property
     def tags(self) -> AsyncTags:
         """Tag management for the dataset.
@@ -4078,10 +4292,7 @@ class AsyncTable:
             be deleted unless they are at least 7 days old. If delete_unverified is True
             then these files will be deleted regardless of their age.
         retrain: bool, default False
-            If True, retrain the vector indices, this would refine the IVF clustering
-            and quantization, which may improve the search accuracy. It's faster than
-            re-creating the index from scratch, so it's recommended to try this first,
-            when the data distribution has changed significantly.
+            This parameter is no longer used and is deprecated.
 
         Experimental API
         ----------------
@@ -4104,10 +4315,19 @@ class AsyncTable:
         cleanup_since_ms: Optional[int] = None
         if cleanup_older_than is not None:
             cleanup_since_ms = round(cleanup_older_than.total_seconds() * 1000)
+
+        if retrain:
+            import warnings
+
+            warnings.warn(
+                "The 'retrain' parameter is deprecated and will be removed in a "
+                "future version.",
+                DeprecationWarning,
+            )
+
         return await self._inner.optimize(
             cleanup_since_ms=cleanup_since_ms,
             delete_unverified=delete_unverified,
-            retrain=retrain,
         )
 
     async def list_indices(self) -> Iterable[IndexConfig]:

python/python/tests/test_db.py

@@ -175,6 +175,18 @@ def test_table_names(tmp_db: lancedb.DBConnection):
     tmp_db.create_table("test3", data=data)
     assert tmp_db.table_names() == ["test1", "test2", "test3"]
 
+    # Test that positional arguments for page_token and limit
+    result = list(tmp_db.table_names("test1", 1))  # page_token="test1", limit=1
+    assert result == ["test2"], f"Expected ['test2'], got {result}"
+
+    # Test mixed positional and keyword arguments
+    result = list(tmp_db.table_names("test2", limit=2))
+    assert result == ["test3"], f"Expected ['test3'], got {result}"
+
+    # Test that namespace parameter can be passed as keyword
+    result = list(tmp_db.table_names(namespace=[]))
+    assert len(result) == 3
+
 
 @pytest.mark.asyncio
 async def test_table_names_async(tmp_path):
@@ -728,3 +740,210 @@ def test_bypass_vector_index_sync(tmp_db: lancedb.DBConnection):
         table.search(sample_key).bypass_vector_index().explain_plan(verbose=True)
     )
     assert "KNN" in plan_without_index
+
+
+def test_local_namespace_operations(tmp_path):
+    """Test that local mode namespace operations behave as expected."""
+    # Create a local database connection
+    db = lancedb.connect(tmp_path)
+
+    # Test list_namespaces returns empty list for root namespace
+    namespaces = list(db.list_namespaces())
+    assert namespaces == []
+
+    # Test list_namespaces with non-empty namespace raises NotImplementedError
+    with pytest.raises(
+        NotImplementedError,
+        match="Namespace operations are not supported for listing database",
+    ):
+        list(db.list_namespaces(namespace=["test"]))
+
+
+def test_local_create_namespace_not_supported(tmp_path):
+    """Test that create_namespace is not supported in local mode."""
+    db = lancedb.connect(tmp_path)
+
+    with pytest.raises(
+        NotImplementedError,
+        match="Namespace operations are not supported for listing database",
+    ):
+        db.create_namespace(["test_namespace"])
+
+
+def test_local_drop_namespace_not_supported(tmp_path):
+    """Test that drop_namespace is not supported in local mode."""
+    db = lancedb.connect(tmp_path)
+
+    with pytest.raises(
+        NotImplementedError,
+        match="Namespace operations are not supported for listing database",
+    ):
+        db.drop_namespace(["test_namespace"])
+
+
+def test_local_table_operations_with_namespace_raise_error(tmp_path):
+    """
+    Test that table operations with namespace parameter
+    raise ValueError in local mode.
+    """
+    db = lancedb.connect(tmp_path)
+
+    # Create some test data
+    data = [{"vector": [1.0, 2.0], "item": "test"}]
+    schema = pa.schema(
+        [pa.field("vector", pa.list_(pa.float32(), 2)), pa.field("item", pa.string())]
+    )
+
+    # Test create_table with namespace - should raise ValueError
+    with pytest.raises(
+        NotImplementedError,
+        match="Namespace parameter is not supported for listing database",
+    ):
+        db.create_table(
+            "test_table_with_ns", data=data, schema=schema, namespace=["test_ns"]
+        )
+
+    # Create table normally for other tests
+    db.create_table("test_table", data=data, schema=schema)
+    assert "test_table" in db.table_names()
+
+    # Test open_table with namespace - should raise ValueError
+    with pytest.raises(
+        NotImplementedError,
+        match="Namespace parameter is not supported for listing database",
+    ):
+        db.open_table("test_table", namespace=["test_ns"])
+
+    # Test table_names with namespace - should raise ValueError
+    with pytest.raises(
+        NotImplementedError,
+        match="Namespace parameter is not supported for listing database",
+    ):
+        list(db.table_names(namespace=["test_ns"]))
+
+    # Test drop_table with namespace - should raise ValueError
+    with pytest.raises(
+        NotImplementedError,
+        match="Namespace parameter is not supported for listing database",
+    ):
+        db.drop_table("test_table", namespace=["test_ns"])
+
+    # Test table_names without namespace - should work normally
+    tables_root = list(db.table_names())
+    assert "test_table" in tables_root
+
+
+def test_clone_table_latest_version(tmp_path):
+    """Test cloning a table with the latest version (default behavior)"""
+    import os
+
+    db = lancedb.connect(tmp_path)
+
+    # Create source table with some data
+    data = [
+        {"id": 1, "text": "hello", "vector": [1.0, 2.0]},
+        {"id": 2, "text": "world", "vector": [3.0, 4.0]},
+    ]
+    source_table = db.create_table("source", data=data)
+
+    # Add more data to create a new version
+    more_data = [{"id": 3, "text": "test", "vector": [5.0, 6.0]}]
+    source_table.add(more_data)
+
+    # Clone the table (should get latest version with 3 rows)
+    source_uri = os.path.join(tmp_path, "source.lance")
+    cloned_table = db.clone_table("cloned", source_uri)
+
+    # Verify cloned table has all 3 rows
+    assert cloned_table.count_rows() == 3
+    assert "cloned" in db.table_names()
+
+    # Verify data matches
+    cloned_data = cloned_table.to_pandas()
+    assert len(cloned_data) == 3
+    assert set(cloned_data["id"].tolist()) == {1, 2, 3}
+
+
+def test_clone_table_specific_version(tmp_path):
+    """Test cloning a table from a specific version"""
+    import os
+
+    db = lancedb.connect(tmp_path)
+
+    # Create source table with initial data
+    data = [
+        {"id": 1, "text": "hello", "vector": [1.0, 2.0]},
+        {"id": 2, "text": "world", "vector": [3.0, 4.0]},
+    ]
+    source_table = db.create_table("source", data=data)
+
+    # Get the initial version
+    initial_version = source_table.version
+
+    # Add more data to create a new version
+    more_data = [{"id": 3, "text": "test", "vector": [5.0, 6.0]}]
+    source_table.add(more_data)
+
+    # Verify source now has 3 rows
+    assert source_table.count_rows() == 3
+
+    # Clone from the initial version (should have only 2 rows)
+    source_uri = os.path.join(tmp_path, "source.lance")
+    cloned_table = db.clone_table("cloned", source_uri, source_version=initial_version)
+
+    # Verify cloned table has only the initial 2 rows
+    assert cloned_table.count_rows() == 2
+    cloned_data = cloned_table.to_pandas()
+    assert set(cloned_data["id"].tolist()) == {1, 2}
+
+
+def test_clone_table_with_tag(tmp_path):
+    """Test cloning a table from a tagged version"""
+    import os
+
+    db = lancedb.connect(tmp_path)
+
+    # Create source table with initial data
+    data = [
+        {"id": 1, "text": "hello", "vector": [1.0, 2.0]},
+        {"id": 2, "text": "world", "vector": [3.0, 4.0]},
+    ]
+    source_table = db.create_table("source", data=data)
+
+    # Create a tag for the current version
+    source_table.tags.create("v1.0", source_table.version)
+
+    # Add more data after the tag
+    more_data = [{"id": 3, "text": "test", "vector": [5.0, 6.0]}]
+    source_table.add(more_data)
+
+    # Verify source now has 3 rows
+    assert source_table.count_rows() == 3
+
+    # Clone from the tagged version (should have only 2 rows)
+    source_uri = os.path.join(tmp_path, "source.lance")
+    cloned_table = db.clone_table("cloned", source_uri, source_tag="v1.0")
+
+    # Verify cloned table has only the tagged version's 2 rows
+    assert cloned_table.count_rows() == 2
+    cloned_data = cloned_table.to_pandas()
+    assert set(cloned_data["id"].tolist()) == {1, 2}
+
+
+def test_clone_table_deep_clone_fails(tmp_path):
+    """Test that deep clone raises an unsupported error"""
+    import os
+
+    db = lancedb.connect(tmp_path)
+
+    # Create source table with some data
+    data = [
+        {"id": 1, "text": "hello", "vector": [1.0, 2.0]},
+        {"id": 2, "text": "world", "vector": [3.0, 4.0]},
+    ]
+    db.create_table("source", data=data)
+
+    # Try to create a deep clone (should fail)
+    source_uri = os.path.join(tmp_path, "source.lance")
+    with pytest.raises(Exception, match="Deep clone is not yet implemented"):
+        db.clone_table("cloned", source_uri, is_shallow=False)

python/python/tests/test_embeddings.py

@@ -114,6 +114,63 @@ def test_embedding_function_variables():
     assert func.safe_model_dump()["secret_key"] == "$var:secret"
 
 
+def test_parse_functions_with_variables():
+    @register("variable-parsing-test")
+    class VariableParsingFunction(TextEmbeddingFunction):
+        api_key: str
+        base_url: Optional[str] = None
+
+        @staticmethod
+        def sensitive_keys():
+            return ["api_key"]
+
+        def ndims(self):
+            return 10
+
+        def generate_embeddings(self, texts):
+            # Mock implementation that just returns random embeddings
+            # In real usage, this would use the api_key to call an API
+            return [np.random.rand(self.ndims()).tolist() for _ in texts]
+
+    registry = EmbeddingFunctionRegistry.get_instance()
+
+    registry.set_var("test_api_key", "sk-test-key-12345")
+    registry.set_var("test_base_url", "https://api.example.com")
+
+    conf = EmbeddingFunctionConfig(
+        source_column="text",
+        vector_column="vector",
+        function=registry.get("variable-parsing-test").create(
+            api_key="$var:test_api_key", base_url="$var:test_base_url"
+        ),
+    )
+
+    metadata = registry.get_table_metadata([conf])
+
+    # Create a mock arrow table with the metadata
+    schema = pa.schema(
+        [pa.field("text", pa.string()), pa.field("vector", pa.list_(pa.float32(), 10))]
+    )
+    table = pa.table({"text": [], "vector": []}, schema=schema)
+    table = table.replace_schema_metadata(metadata)
+
+    ds = lance.write_dataset(table, "memory://")
+
+    configs = registry.parse_functions(ds.schema.metadata)
+
+    assert "vector" in configs
+    parsed_func = configs["vector"].function
+
+    assert parsed_func.api_key == "sk-test-key-12345"
+    assert parsed_func.base_url == "https://api.example.com"
+
+    embeddings = parsed_func.generate_embeddings(["test text"])
+    assert len(embeddings) == 1
+    assert len(embeddings[0]) == 10
+
+    assert parsed_func.safe_model_dump()["api_key"] == "$var:test_api_key"
+
+
 def test_embedding_with_bad_results(tmp_path):
     @register("null-embedding")
     class NullEmbeddingFunction(TextEmbeddingFunction):

python/python/tests/test_fts.py

@@ -157,7 +157,16 @@ def test_create_index_with_stemming(tmp_path, table):
 def test_create_inverted_index(table, use_tantivy, with_position):
     if use_tantivy and not with_position:
         pytest.skip("we don't support building a tantivy index without position")
-    table.create_fts_index("text", use_tantivy=use_tantivy, with_position=with_position)
+    table.create_fts_index(
+        "text",
+        use_tantivy=use_tantivy,
+        with_position=with_position,
+        name="custom_fts_index",
+    )
+    if not use_tantivy:
+        indices = table.list_indices()
+        fts_indices = [i for i in indices if i.index_type == "FTS"]
+        assert any(i.name == "custom_fts_index" for i in fts_indices)
 
 
 def test_populate_index(tmp_path, table):

python/python/tests/test_header_provider.py

@@ -0,0 +1,237 @@
+# SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright The LanceDB Authors
+
+import concurrent.futures
+import pytest
+import time
+import threading
+from typing import Dict
+
+from lancedb.remote import ClientConfig, HeaderProvider
+from lancedb.remote.header import StaticHeaderProvider, OAuthProvider
+
+
+class TestStaticHeaderProvider:
+    def test_init(self):
+        """Test StaticHeaderProvider initialization."""
+        headers = {"X-API-Key": "test-key", "X-Custom": "value"}
+        provider = StaticHeaderProvider(headers)
+        assert provider._headers == headers
+
+    def test_get_headers(self):
+        """Test get_headers returns correct headers."""
+        headers = {"X-API-Key": "test-key", "X-Custom": "value"}
+        provider = StaticHeaderProvider(headers)
+
+        result = provider.get_headers()
+        assert result == headers
+
+        # Ensure it returns a copy
+        result["X-Modified"] = "modified"
+        result2 = provider.get_headers()
+        assert "X-Modified" not in result2
+
+
+class TestOAuthProvider:
+    def test_init(self):
+        """Test OAuthProvider initialization."""
+
+        def fetcher():
+            return {"access_token": "token123", "expires_in": 3600}
+
+        provider = OAuthProvider(fetcher)
+        assert provider._token_fetcher is fetcher
+        assert provider._refresh_buffer == 300
+        assert provider._current_token is None
+        assert provider._token_expires_at is None
+
+    def test_get_headers_first_time(self):
+        """Test get_headers fetches token on first call."""
+
+        def fetcher():
+            return {"access_token": "token123", "expires_in": 3600}
+
+        provider = OAuthProvider(fetcher)
+        headers = provider.get_headers()
+
+        assert headers == {"Authorization": "Bearer token123"}
+        assert provider._current_token == "token123"
+        assert provider._token_expires_at is not None
+
+    def test_token_refresh(self):
+        """Test token refresh when expired."""
+        call_count = 0
+        tokens = ["token1", "token2"]
+
+        def fetcher():
+            nonlocal call_count
+            token = tokens[call_count]
+            call_count += 1
+            return {"access_token": token, "expires_in": 1}  # Expires in 1 second
+
+        provider = OAuthProvider(fetcher, refresh_buffer_seconds=0)
+
+        # First call
+        headers1 = provider.get_headers()
+        assert headers1 == {"Authorization": "Bearer token1"}
+
+        # Wait for token to expire
+        time.sleep(1.1)
+
+        # Second call should refresh
+        headers2 = provider.get_headers()
+        assert headers2 == {"Authorization": "Bearer token2"}
+        assert call_count == 2
+
+    def test_no_expiry_info(self):
+        """Test handling tokens without expiry information."""
+
+        def fetcher():
+            return {"access_token": "permanent_token"}
+
+        provider = OAuthProvider(fetcher)
+        headers = provider.get_headers()
+
+        assert headers == {"Authorization": "Bearer permanent_token"}
+        assert provider._token_expires_at is None
+
+        # Should not refresh on second call
+        headers2 = provider.get_headers()
+        assert headers2 == {"Authorization": "Bearer permanent_token"}
+
+    def test_missing_access_token(self):
+        """Test error handling when access_token is missing."""
+
+        def fetcher():
+            return {"expires_in": 3600}  # Missing access_token
+
+        provider = OAuthProvider(fetcher)
+
+        with pytest.raises(
+            ValueError, match="Token fetcher did not return 'access_token'"
+        ):
+            provider.get_headers()
+
+    def test_sync_method(self):
+        """Test synchronous get_headers method."""
+
+        def fetcher():
+            return {"access_token": "sync_token", "expires_in": 3600}
+
+        provider = OAuthProvider(fetcher)
+        headers = provider.get_headers()
+
+        assert headers == {"Authorization": "Bearer sync_token"}
+
+
+class TestClientConfigIntegration:
+    def test_client_config_with_header_provider(self):
+        """Test ClientConfig can accept a HeaderProvider."""
+        provider = StaticHeaderProvider({"X-Test": "value"})
+        config = ClientConfig(header_provider=provider)
+
+        assert config.header_provider is provider
+
+    def test_client_config_without_header_provider(self):
+        """Test ClientConfig works without HeaderProvider."""
+        config = ClientConfig()
+        assert config.header_provider is None
+
+
+class CustomProvider(HeaderProvider):
+    """Custom provider for testing abstract class."""
+
+    def get_headers(self) -> Dict[str, str]:
+        return {"X-Custom": "custom-value"}
+
+
+class TestCustomHeaderProvider:
+    def test_custom_provider(self):
+        """Test custom HeaderProvider implementation."""
+        provider = CustomProvider()
+        headers = provider.get_headers()
+        assert headers == {"X-Custom": "custom-value"}
+
+
+class ErrorProvider(HeaderProvider):
+    """Provider that raises errors for testing error handling."""
+
+    def __init__(self, error_message: str = "Test error"):
+        self.error_message = error_message
+        self.call_count = 0
+
+    def get_headers(self) -> Dict[str, str]:
+        self.call_count += 1
+        raise RuntimeError(self.error_message)
+
+
+class TestErrorHandling:
+    def test_provider_error_propagation(self):
+        """Test that errors from header provider are properly propagated."""
+        provider = ErrorProvider("Authentication failed")
+
+        with pytest.raises(RuntimeError, match="Authentication failed"):
+            provider.get_headers()
+
+        assert provider.call_count == 1
+
+    def test_provider_error(self):
+        """Test that errors are propagated."""
+        provider = ErrorProvider("Sync error")
+
+        with pytest.raises(RuntimeError, match="Sync error"):
+            provider.get_headers()
+
+
+class ConcurrentProvider(HeaderProvider):
+    """Provider for testing thread safety."""
+
+    def __init__(self):
+        self.counter = 0
+        self.lock = threading.Lock()
+
+    def get_headers(self) -> Dict[str, str]:
+        with self.lock:
+            self.counter += 1
+            # Simulate some work
+            time.sleep(0.01)
+            return {"X-Request-Id": str(self.counter)}
+
+
+class TestConcurrency:
+    def test_concurrent_header_fetches(self):
+        """Test that header provider can handle concurrent requests."""
+        provider = ConcurrentProvider()
+
+        # Create multiple concurrent requests
+        with concurrent.futures.ThreadPoolExecutor(max_workers=10) as executor:
+            futures = [executor.submit(provider.get_headers) for _ in range(10)]
+            results = [f.result() for f in futures]
+
+        # Each request should get a unique counter value
+        request_ids = [int(r["X-Request-Id"]) for r in results]
+        assert len(set(request_ids)) == 10
+        assert min(request_ids) == 1
+        assert max(request_ids) == 10
+
+    def test_oauth_concurrent_refresh(self):
+        """Test that OAuth provider handles concurrent refresh requests safely."""
+        call_count = 0
+
+        def slow_token_fetch():
+            nonlocal call_count
+            call_count += 1
+            time.sleep(0.1)  # Simulate slow token fetch
+            return {"access_token": f"token-{call_count}", "expires_in": 3600}
+
+        provider = OAuthProvider(slow_token_fetch)
+
+        # Force multiple concurrent refreshes
+        with concurrent.futures.ThreadPoolExecutor(max_workers=5) as executor:
+            futures = [executor.submit(provider.get_headers) for _ in range(5)]
+            results = [f.result() for f in futures]
+
+        # All requests should get the same token (only one refresh should happen)
+        tokens = [r["Authorization"] for r in results]
+        assert all(t == "Bearer token-1" for t in tokens)
+        assert call_count == 1  # Only one token fetch despite concurrent requests

python/python/tests/test_namespace.py

@@ -0,0 +1,707 @@
+# SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright The LanceDB Authors
+
+"""Tests for LanceDB namespace integration."""
+
+import tempfile
+import shutil
+from typing import Dict, Optional
+import pytest
+import pyarrow as pa
+import lancedb
+from lance_namespace.namespace import NATIVE_IMPLS, LanceNamespace
+from lance_namespace_urllib3_client.models import (
+    ListTablesRequest,
+    ListTablesResponse,
+    DescribeTableRequest,
+    DescribeTableResponse,
+    RegisterTableRequest,
+    RegisterTableResponse,
+    DeregisterTableRequest,
+    DeregisterTableResponse,
+    CreateTableRequest,
+    CreateTableResponse,
+    DropTableRequest,
+    DropTableResponse,
+    ListNamespacesRequest,
+    ListNamespacesResponse,
+    CreateNamespaceRequest,
+    CreateNamespaceResponse,
+    DropNamespaceRequest,
+    DropNamespaceResponse,
+)
+
+
+class TempNamespace(LanceNamespace):
+    """A simple dictionary-backed namespace for testing."""
+
+    # Class-level storage to persist table registry across instances
+    _global_registry: Dict[str, Dict[str, str]] = {}
+    # Class-level storage for namespaces (supporting 1-level namespace)
+    _global_namespaces: Dict[str, set] = {}
+
+    def __init__(self, **properties):
+        """Initialize the test namespace.
+
+        Args:
+            root: The root directory for tables (optional)
+            **properties: Additional configuration properties
+        """
+        self.config = TempNamespaceConfig(properties)
+        # Use the root as a key to maintain separate registries per root
+        root = self.config.root
+        if root not in self._global_registry:
+            self._global_registry[root] = {}
+        if root not in self._global_namespaces:
+            self._global_namespaces[root] = set()
+        self.tables = self._global_registry[root]  # Reference to shared registry
+        self.namespaces = self._global_namespaces[
+            root
+        ]  # Reference to shared namespaces
+
+    def list_tables(self, request: ListTablesRequest) -> ListTablesResponse:
+        """List all tables in the namespace."""
+        if not request.id:
+            # List all tables in root namespace
+            tables = [name for name in self.tables.keys() if "." not in name]
+        else:
+            # List tables in specific namespace (1-level only)
+            if len(request.id) == 1:
+                namespace_name = request.id[0]
+                prefix = f"{namespace_name}."
+                tables = [
+                    name[len(prefix) :]
+                    for name in self.tables.keys()
+                    if name.startswith(prefix)
+                ]
+            else:
+                # Multi-level namespaces not supported
+                raise ValueError("Only 1-level namespaces are supported")
+        return ListTablesResponse(tables=tables)
+
+    def describe_table(self, request: DescribeTableRequest) -> DescribeTableResponse:
+        """Describe a table by returning its location."""
+        if not request.id:
+            raise ValueError("Invalid table ID")
+
+        if len(request.id) == 1:
+            # Root namespace table
+            table_name = request.id[0]
+        elif len(request.id) == 2:
+            # Namespaced table (1-level namespace)
+            namespace_name, table_name = request.id
+            table_name = f"{namespace_name}.{table_name}"
+        else:
+            raise ValueError("Only 1-level namespaces are supported")
+
+        if table_name not in self.tables:
+            raise RuntimeError(f"Table does not exist: {table_name}")
+
+        table_uri = self.tables[table_name]
+        return DescribeTableResponse(location=table_uri)
+
+    def create_table(
+        self, request: CreateTableRequest, request_data: bytes
+    ) -> CreateTableResponse:
+        """Create a table in the namespace."""
+        if not request.id:
+            raise ValueError("Invalid table ID")
+
+        if len(request.id) == 1:
+            # Root namespace table
+            table_name = request.id[0]
+            table_uri = f"{self.config.root}/{table_name}.lance"
+        elif len(request.id) == 2:
+            # Namespaced table (1-level namespace)
+            namespace_name, base_table_name = request.id
+            # Add namespace to our namespace set
+            self.namespaces.add(namespace_name)
+            table_name = f"{namespace_name}.{base_table_name}"
+            table_uri = f"{self.config.root}/{namespace_name}/{base_table_name}.lance"
+        else:
+            raise ValueError("Only 1-level namespaces are supported")
+
+        # Check if table already exists
+        if table_name in self.tables:
+            if request.mode == "overwrite":
+                # Drop existing table for overwrite mode
+                del self.tables[table_name]
+            else:
+                raise RuntimeError(f"Table already exists: {table_name}")
+
+        # Parse the Arrow IPC stream to get the schema and create the actual table
+        import pyarrow.ipc as ipc
+        import io
+        import lance
+        import os
+
+        # Create directory if needed for namespaced tables
+        os.makedirs(os.path.dirname(table_uri), exist_ok=True)
+
+        # Read the IPC stream
+        reader = ipc.open_stream(io.BytesIO(request_data))
+        table = reader.read_all()
+
+        # Create the actual Lance table
+        lance.write_dataset(table, table_uri)
+
+        # Store the table mapping
+        self.tables[table_name] = table_uri
+
+        return CreateTableResponse(location=table_uri)
+
+    def drop_table(self, request: DropTableRequest) -> DropTableResponse:
+        """Drop a table from the namespace."""
+        if not request.id:
+            raise ValueError("Invalid table ID")
+
+        if len(request.id) == 1:
+            # Root namespace table
+            table_name = request.id[0]
+        elif len(request.id) == 2:
+            # Namespaced table (1-level namespace)
+            namespace_name, base_table_name = request.id
+            table_name = f"{namespace_name}.{base_table_name}"
+        else:
+            raise ValueError("Only 1-level namespaces are supported")
+
+        if table_name not in self.tables:
+            raise RuntimeError(f"Table does not exist: {table_name}")
+
+        # Get the table URI
+        table_uri = self.tables[table_name]
+
+        # Delete the actual table files
+        import shutil
+        import os
+
+        if os.path.exists(table_uri):
+            shutil.rmtree(table_uri, ignore_errors=True)
+
+        # Remove from registry
+        del self.tables[table_name]
+
+        return DropTableResponse()
+
+    def register_table(self, request: RegisterTableRequest) -> RegisterTableResponse:
+        """Register a table with the namespace."""
+        if not request.id or len(request.id) != 1:
+            raise ValueError("Invalid table ID")
+
+        if not request.location:
+            raise ValueError("Table location is required")
+
+        table_name = request.id[0]
+        self.tables[table_name] = request.location
+
+        return RegisterTableResponse()
+
+    def deregister_table(
+        self, request: DeregisterTableRequest
+    ) -> DeregisterTableResponse:
+        """Deregister a table from the namespace."""
+        if not request.id or len(request.id) != 1:
+            raise ValueError("Invalid table ID")
+
+        table_name = request.id[0]
+        if table_name not in self.tables:
+            raise RuntimeError(f"Table does not exist: {table_name}")
+
+        del self.tables[table_name]
+        return DeregisterTableResponse()
+
+    def list_namespaces(self, request: ListNamespacesRequest) -> ListNamespacesResponse:
+        """List child namespaces."""
+        if not request.id:
+            # List root-level namespaces
+            namespaces = list(self.namespaces)
+        elif len(request.id) == 1:
+            # For 1-level namespace, there are no child namespaces
+            namespaces = []
+        else:
+            raise ValueError("Only 1-level namespaces are supported")
+
+        return ListNamespacesResponse(namespaces=namespaces)
+
+    def create_namespace(
+        self, request: CreateNamespaceRequest
+    ) -> CreateNamespaceResponse:
+        """Create a namespace."""
+        if not request.id:
+            raise ValueError("Invalid namespace ID")
+
+        if len(request.id) == 1:
+            # Create 1-level namespace
+            namespace_name = request.id[0]
+            self.namespaces.add(namespace_name)
+
+            # Create directory for the namespace
+            import os
+
+            namespace_dir = f"{self.config.root}/{namespace_name}"
+            os.makedirs(namespace_dir, exist_ok=True)
+        else:
+            raise ValueError("Only 1-level namespaces are supported")
+
+        return CreateNamespaceResponse()
+
+    def drop_namespace(self, request: DropNamespaceRequest) -> DropNamespaceResponse:
+        """Drop a namespace."""
+        if not request.id:
+            raise ValueError("Invalid namespace ID")
+
+        if len(request.id) == 1:
+            # Drop 1-level namespace
+            namespace_name = request.id[0]
+
+            if namespace_name not in self.namespaces:
+                raise RuntimeError(f"Namespace does not exist: {namespace_name}")
+
+            # Check if namespace has any tables
+            prefix = f"{namespace_name}."
+            tables_in_namespace = [
+                name for name in self.tables.keys() if name.startswith(prefix)
+            ]
+            if tables_in_namespace:
+                raise RuntimeError(
+                    f"Cannot drop namespace '{namespace_name}': contains tables"
+                )
+
+            # Remove namespace
+            self.namespaces.remove(namespace_name)
+
+            # Remove directory
+            import shutil
+            import os
+
+            namespace_dir = f"{self.config.root}/{namespace_name}"
+            if os.path.exists(namespace_dir):
+                shutil.rmtree(namespace_dir, ignore_errors=True)
+        else:
+            raise ValueError("Only 1-level namespaces are supported")
+
+        return DropNamespaceResponse()
+
+
+class TempNamespaceConfig:
+    """Configuration for TestNamespace."""
+
+    ROOT = "root"
+
+    def __init__(self, properties: Optional[Dict[str, str]] = None):
+        """Initialize configuration from properties.
+
+        Args:
+            properties: Dictionary of configuration properties
+        """
+        if properties is None:
+            properties = {}
+
+        self._root = properties.get(self.ROOT, "/tmp")
+
+    @property
+    def root(self) -> str:
+        """Get the namespace root directory."""
+        return self._root
+
+
+NATIVE_IMPLS["temp"] = f"{TempNamespace.__module__}.TempNamespace"
+
+
+class TestNamespaceConnection:
+    """Test namespace-based LanceDB connection."""
+
+    def setup_method(self):
+        """Set up test fixtures."""
+        self.temp_dir = tempfile.mkdtemp()
+        # Clear the TestNamespace registry for this test
+        if self.temp_dir in TempNamespace._global_registry:
+            TempNamespace._global_registry[self.temp_dir].clear()
+        if self.temp_dir in TempNamespace._global_namespaces:
+            TempNamespace._global_namespaces[self.temp_dir].clear()
+
+    def teardown_method(self):
+        """Clean up test fixtures."""
+        # Clear the TestNamespace registry
+        if self.temp_dir in TempNamespace._global_registry:
+            del TempNamespace._global_registry[self.temp_dir]
+        if self.temp_dir in TempNamespace._global_namespaces:
+            del TempNamespace._global_namespaces[self.temp_dir]
+        shutil.rmtree(self.temp_dir, ignore_errors=True)
+
+    def test_connect_namespace_test(self):
+        """Test connecting to LanceDB through TestNamespace."""
+        # Connect using TestNamespace
+        db = lancedb.connect_namespace("temp", {"root": self.temp_dir})
+
+        # Should be a LanceNamespaceDBConnection
+        assert isinstance(db, lancedb.LanceNamespaceDBConnection)
+
+        # Initially no tables
+        assert len(list(db.table_names())) == 0
+
+    def test_create_table_through_namespace(self):
+        """Test creating a table through namespace."""
+        db = lancedb.connect_namespace("temp", {"root": self.temp_dir})
+
+        # Define schema for empty table
+        schema = pa.schema(
+            [
+                pa.field("id", pa.int64()),
+                pa.field("vector", pa.list_(pa.float32(), 2)),
+                pa.field("text", pa.string()),
+            ]
+        )
+
+        # Create empty table
+        table = db.create_table("test_table", schema=schema)
+        assert table is not None
+        assert table.name == "test_table"
+
+        # Table should appear in namespace
+        table_names = list(db.table_names())
+        assert "test_table" in table_names
+        assert len(table_names) == 1
+
+        # Verify empty table
+        result = table.to_pandas()
+        assert len(result) == 0
+        assert list(result.columns) == ["id", "vector", "text"]
+
+    def test_open_table_through_namespace(self):
+        """Test opening an existing table through namespace."""
+        db = lancedb.connect_namespace("temp", {"root": self.temp_dir})
+
+        # Create a table with schema
+        schema = pa.schema(
+            [
+                pa.field("id", pa.int64()),
+                pa.field("vector", pa.list_(pa.float32(), 2)),
+            ]
+        )
+        db.create_table("test_table", schema=schema)
+
+        # Open the table
+        table = db.open_table("test_table")
+        assert table is not None
+        assert table.name == "test_table"
+
+        # Verify empty table with correct schema
+        result = table.to_pandas()
+        assert len(result) == 0
+        assert list(result.columns) == ["id", "vector"]
+
+    def test_drop_table_through_namespace(self):
+        """Test dropping a table through namespace."""
+        db = lancedb.connect_namespace("temp", {"root": self.temp_dir})
+
+        # Create tables
+        schema = pa.schema(
+            [
+                pa.field("id", pa.int64()),
+                pa.field("vector", pa.list_(pa.float32(), 2)),
+            ]
+        )
+        db.create_table("table1", schema=schema)
+        db.create_table("table2", schema=schema)
+
+        # Verify both tables exist
+        table_names = list(db.table_names())
+        assert "table1" in table_names
+        assert "table2" in table_names
+        assert len(table_names) == 2
+
+        # Drop one table
+        db.drop_table("table1")
+
+        # Verify only table2 remains
+        table_names = list(db.table_names())
+        assert "table1" not in table_names
+        assert "table2" in table_names
+        assert len(table_names) == 1
+
+        # Test that drop_table works without explicit namespace parameter
+        db.drop_table("table2")
+        assert len(list(db.table_names())) == 0
+
+        # Should not be able to open dropped table
+        with pytest.raises(RuntimeError):
+            db.open_table("table1")
+
+    def test_create_table_with_schema(self):
+        """Test creating a table with explicit schema through namespace."""
+        db = lancedb.connect_namespace("temp", {"root": self.temp_dir})
+
+        # Define schema
+        schema = pa.schema(
+            [
+                pa.field("id", pa.int64()),
+                pa.field("vector", pa.list_(pa.float32(), 3)),
+                pa.field("text", pa.string()),
+            ]
+        )
+
+        # Create table with schema
+        table = db.create_table("test_table", schema=schema)
+        assert table is not None
+
+        # Verify schema
+        table_schema = table.schema
+        assert len(table_schema) == 3
+        assert table_schema.field("id").type == pa.int64()
+        assert table_schema.field("text").type == pa.string()
+
+    def test_rename_table_not_supported(self):
+        """Test that rename_table raises NotImplementedError."""
+        db = lancedb.connect_namespace("temp", {"root": self.temp_dir})
+
+        # Create a table
+        schema = pa.schema(
+            [
+                pa.field("id", pa.int64()),
+                pa.field("vector", pa.list_(pa.float32(), 2)),
+            ]
+        )
+        db.create_table("old_name", schema=schema)
+
+        # Rename should raise NotImplementedError
+        with pytest.raises(NotImplementedError, match="rename_table is not supported"):
+            db.rename_table("old_name", "new_name")
+
+    def test_drop_all_tables(self):
+        """Test dropping all tables through namespace."""
+        db = lancedb.connect_namespace("temp", {"root": self.temp_dir})
+
+        # Create multiple tables
+        schema = pa.schema(
+            [
+                pa.field("id", pa.int64()),
+                pa.field("vector", pa.list_(pa.float32(), 2)),
+            ]
+        )
+        for i in range(3):
+            db.create_table(f"table{i}", schema=schema)
+
+        # Verify tables exist
+        assert len(list(db.table_names())) == 3
+
+        # Drop all tables
+        db.drop_all_tables()
+
+        # Verify all tables are gone
+        assert len(list(db.table_names())) == 0
+
+        # Test that table_names works with keyword-only namespace parameter
+        db.create_table("test_table", schema=schema)
+        result = list(db.table_names(namespace=[]))
+        assert "test_table" in result
+
+    def test_table_operations(self):
+        """Test various table operations through namespace."""
+        db = lancedb.connect_namespace("temp", {"root": self.temp_dir})
+
+        # Create a table with schema
+        schema = pa.schema(
+            [
+                pa.field("id", pa.int64()),
+                pa.field("vector", pa.list_(pa.float32(), 2)),
+                pa.field("text", pa.string()),
+            ]
+        )
+        table = db.create_table("test_table", schema=schema)
+
+        # Verify empty table was created
+        result = table.to_pandas()
+        assert len(result) == 0
+        assert list(result.columns) == ["id", "vector", "text"]
+
+        # Test add data to the table
+        new_data = [
+            {"id": 1, "vector": [1.0, 2.0], "text": "item_1"},
+            {"id": 2, "vector": [2.0, 3.0], "text": "item_2"},
+        ]
+        table.add(new_data)
+        result = table.to_pandas()
+        assert len(result) == 2
+
+        # Test delete
+        table.delete("id = 1")
+        result = table.to_pandas()
+        assert len(result) == 1
+        assert result["id"].values[0] == 2
+
+        # Test update
+        table.update(where="id = 2", values={"text": "updated"})
+        result = table.to_pandas()
+        assert result["text"].values[0] == "updated"
+
+    def test_storage_options(self):
+        """Test passing storage options through namespace connection."""
+        # Connect with storage options
+        storage_opts = {"test_option": "test_value"}
+        db = lancedb.connect_namespace(
+            "temp", {"root": self.temp_dir}, storage_options=storage_opts
+        )
+
+        # Storage options should be preserved
+        assert db.storage_options == storage_opts
+
+        # Create table with additional storage options
+        table_opts = {"table_option": "table_value"}
+        schema = pa.schema(
+            [
+                pa.field("id", pa.int64()),
+                pa.field("vector", pa.list_(pa.float32(), 2)),
+            ]
+        )
+        db.create_table("test_table", schema=schema, storage_options=table_opts)
+
+    def test_namespace_operations(self):
+        """Test namespace management operations."""
+        db = lancedb.connect_namespace("temp", {"root": self.temp_dir})
+
+        # Initially no namespaces
+        assert len(list(db.list_namespaces())) == 0
+
+        # Create a namespace
+        db.create_namespace(["test_namespace"])
+
+        # Verify namespace exists
+        namespaces = list(db.list_namespaces())
+        assert "test_namespace" in namespaces
+        assert len(namespaces) == 1
+
+        # Create table in namespace
+        schema = pa.schema(
+            [
+                pa.field("id", pa.int64()),
+                pa.field("vector", pa.list_(pa.float32(), 2)),
+            ]
+        )
+        table = db.create_table(
+            "test_table", schema=schema, namespace=["test_namespace"]
+        )
+        assert table is not None
+
+        # Verify table exists in namespace
+        tables_in_namespace = list(db.table_names(namespace=["test_namespace"]))
+        assert "test_table" in tables_in_namespace
+        assert len(tables_in_namespace) == 1
+
+        # Open table from namespace
+        table = db.open_table("test_table", namespace=["test_namespace"])
+        assert table is not None
+        assert table.name == "test_table"
+
+        # Drop table from namespace
+        db.drop_table("test_table", namespace=["test_namespace"])
+
+        # Verify table no longer exists in namespace
+        tables_in_namespace = list(db.table_names(namespace=["test_namespace"]))
+        assert len(tables_in_namespace) == 0
+
+        # Drop namespace
+        db.drop_namespace(["test_namespace"])
+
+        # Verify namespace no longer exists
+        namespaces = list(db.list_namespaces())
+        assert len(namespaces) == 0
+
+    def test_namespace_with_tables_cannot_be_dropped(self):
+        """Test that namespaces containing tables cannot be dropped."""
+        db = lancedb.connect_namespace("temp", {"root": self.temp_dir})
+
+        # Create namespace and table
+        db.create_namespace(["test_namespace"])
+        schema = pa.schema(
+            [
+                pa.field("id", pa.int64()),
+                pa.field("vector", pa.list_(pa.float32(), 2)),
+            ]
+        )
+        db.create_table("test_table", schema=schema, namespace=["test_namespace"])
+
+        # Try to drop namespace with tables - should fail
+        with pytest.raises(RuntimeError, match="contains tables"):
+            db.drop_namespace(["test_namespace"])
+
+        # Drop table first
+        db.drop_table("test_table", namespace=["test_namespace"])
+
+        # Now dropping namespace should work
+        db.drop_namespace(["test_namespace"])
+
+    def test_same_table_name_different_namespaces(self):
+        db = lancedb.connect_namespace("temp", {"root": self.temp_dir})
+
+        # Create two namespaces
+        db.create_namespace(["namespace_a"])
+        db.create_namespace(["namespace_b"])
+
+        # Define schema
+        schema = pa.schema(
+            [
+                pa.field("id", pa.int64()),
+                pa.field("vector", pa.list_(pa.float32(), 2)),
+                pa.field("text", pa.string()),
+            ]
+        )
+
+        # Create table with same name in both namespaces
+        table_a = db.create_table(
+            "same_name_table", schema=schema, namespace=["namespace_a"]
+        )
+        table_b = db.create_table(
+            "same_name_table", schema=schema, namespace=["namespace_b"]
+        )
+
+        # Add different data to each table
+        data_a = [
+            {"id": 1, "vector": [1.0, 2.0], "text": "data_from_namespace_a"},
+            {"id": 2, "vector": [3.0, 4.0], "text": "also_from_namespace_a"},
+        ]
+        table_a.add(data_a)
+
+        data_b = [
+            {"id": 10, "vector": [10.0, 20.0], "text": "data_from_namespace_b"},
+            {"id": 20, "vector": [30.0, 40.0], "text": "also_from_namespace_b"},
+            {"id": 30, "vector": [50.0, 60.0], "text": "more_from_namespace_b"},
+        ]
+        table_b.add(data_b)
+
+        # Verify data in namespace_a table
+        opened_table_a = db.open_table("same_name_table", namespace=["namespace_a"])
+        result_a = opened_table_a.to_pandas().sort_values("id").reset_index(drop=True)
+        assert len(result_a) == 2
+        assert result_a["id"].tolist() == [1, 2]
+        assert result_a["text"].tolist() == [
+            "data_from_namespace_a",
+            "also_from_namespace_a",
+        ]
+        assert [v.tolist() for v in result_a["vector"]] == [[1.0, 2.0], [3.0, 4.0]]
+
+        # Verify data in namespace_b table
+        opened_table_b = db.open_table("same_name_table", namespace=["namespace_b"])
+        result_b = opened_table_b.to_pandas().sort_values("id").reset_index(drop=True)
+        assert len(result_b) == 3
+        assert result_b["id"].tolist() == [10, 20, 30]
+        assert result_b["text"].tolist() == [
+            "data_from_namespace_b",
+            "also_from_namespace_b",
+            "more_from_namespace_b",
+        ]
+        assert [v.tolist() for v in result_b["vector"]] == [
+            [10.0, 20.0],
+            [30.0, 40.0],
+            [50.0, 60.0],
+        ]
+
+        # Verify root namespace doesn't have this table
+        root_tables = list(db.table_names())
+        assert "same_name_table" not in root_tables
+
+        # Clean up
+        db.drop_table("same_name_table", namespace=["namespace_a"])
+        db.drop_table("same_name_table", namespace=["namespace_b"])
+        db.drop_namespace(["namespace_a"])
+        db.drop_namespace(["namespace_b"])

python/python/tests/test_query.py

@@ -5,6 +5,7 @@ from typing import List, Union
 import unittest.mock as mock
 from datetime import timedelta
 from pathlib import Path
+import random
 
 import lancedb
 from lancedb.db import AsyncConnection
@@ -1327,6 +1328,55 @@ def test_query_timeout(tmp_path):
         )
 
 
+def test_take_queries(tmp_path):
+    db = lancedb.connect(tmp_path)
+    data = pa.table(
+        {
+            "idx": range(100),
+        }
+    )
+    table = db.create_table("test", data)
+
+    # Take by offset
+    assert list(
+        sorted(table.take_offsets([5, 2, 17]).to_pandas()["idx"].to_list())
+    ) == [
+        2,
+        5,
+        17,
+    ]
+
+    # Take by row id
+    assert list(
+        sorted(table.take_row_ids([5, 2, 17]).to_pandas()["idx"].to_list())
+    ) == [
+        2,
+        5,
+        17,
+    ]
+
+
+def test_getitems(tmp_path):
+    db = lancedb.connect(tmp_path)
+    data = pa.table(
+        {
+            "idx": range(100),
+        }
+    )
+    # Make two fragments
+    table = db.create_table("test", data)
+    table.add(pa.table({"idx": range(100, 200)}))
+
+    assert table.__getitems__([5, 2, 117]) == pa.table(
+        {
+            "idx": [5, 2, 117],
+        }
+    )
+
+    offsets = random.sample(range(200), 10)
+    assert table.__getitems__(offsets) == pa.table({"idx": offsets})
+
+
 @pytest.mark.asyncio
 async def test_query_timeout_async(tmp_path):
     db = await lancedb.connect_async(tmp_path)

python/python/tests/test_remote_db.py

@@ -7,6 +7,7 @@ from datetime import timedelta
 import http.server
 import json
 import threading
+import time
 from unittest.mock import MagicMock
 import uuid
 from packaging.version import Version
@@ -271,12 +272,21 @@ def test_table_add_in_threadpool():
 
 
 def test_table_create_indices():
+    # Track received index creation requests to validate name parameter
+    received_requests = []
+
     def handler(request):
         index_stats = dict(
             index_type="IVF_PQ", num_indexed_rows=1000, num_unindexed_rows=0
         )
 
         if request.path == "/v1/table/test/create_index/":
+            # Capture the request body to validate name parameter
+            content_len = int(request.headers.get("Content-Length", 0))
+            if content_len > 0:
+                body = request.rfile.read(content_len)
+                body_data = json.loads(body)
+                received_requests.append(body_data)
             request.send_response(200)
             request.end_headers()
         elif request.path == "/v1/table/test/create/?mode=create":
@@ -307,34 +317,34 @@ def test_table_create_indices():
                 dict(
                     indexes=[
                         {
-                            "index_name": "id_idx",
+                            "index_name": "custom_scalar_idx",
                             "columns": ["id"],
                         },
                         {
-                            "index_name": "text_idx",
+                            "index_name": "custom_fts_idx",
                             "columns": ["text"],
                         },
                         {
-                            "index_name": "vector_idx",
+                            "index_name": "custom_vector_idx",
                             "columns": ["vector"],
                         },
                     ]
                 )
             )
             request.wfile.write(payload.encode())
-        elif request.path == "/v1/table/test/index/id_idx/stats/":
+        elif request.path == "/v1/table/test/index/custom_scalar_idx/stats/":
             request.send_response(200)
             request.send_header("Content-Type", "application/json")
             request.end_headers()
             payload = json.dumps(index_stats)
             request.wfile.write(payload.encode())
-        elif request.path == "/v1/table/test/index/text_idx/stats/":
+        elif request.path == "/v1/table/test/index/custom_fts_idx/stats/":
             request.send_response(200)
             request.send_header("Content-Type", "application/json")
             request.end_headers()
             payload = json.dumps(index_stats)
             request.wfile.write(payload.encode())
-        elif request.path == "/v1/table/test/index/vector_idx/stats/":
+        elif request.path == "/v1/table/test/index/custom_vector_idx/stats/":
             request.send_response(200)
             request.send_header("Content-Type", "application/json")
             request.end_headers()
@@ -351,16 +361,49 @@ def test_table_create_indices():
         # Parameters are well-tested through local and async tests.
         # This is a smoke-test.
         table = db.create_table("test", [{"id": 1}])
-        table.create_scalar_index("id", wait_timeout=timedelta(seconds=2))
-        table.create_fts_index("text", wait_timeout=timedelta(seconds=2))
-        table.create_index(
-            vector_column_name="vector", wait_timeout=timedelta(seconds=10)
+
+        # Test create_scalar_index with custom name
+        table.create_scalar_index(
+            "id", wait_timeout=timedelta(seconds=2), name="custom_scalar_idx"
         )
-        table.wait_for_index(["id_idx"], timedelta(seconds=2))
-        table.wait_for_index(["text_idx", "vector_idx"], timedelta(seconds=2))
-        table.drop_index("vector_idx")
-        table.drop_index("id_idx")
-        table.drop_index("text_idx")
+
+        # Test create_fts_index with custom name
+        table.create_fts_index(
+            "text", wait_timeout=timedelta(seconds=2), name="custom_fts_idx"
+        )
+
+        # Test create_index with custom name
+        table.create_index(
+            vector_column_name="vector",
+            wait_timeout=timedelta(seconds=10),
+            name="custom_vector_idx",
+        )
+
+        # Validate that the name parameter was passed correctly in requests
+        assert len(received_requests) == 3
+
+        # Check scalar index request has custom name
+        scalar_req = received_requests[0]
+        assert "name" in scalar_req
+        assert scalar_req["name"] == "custom_scalar_idx"
+
+        # Check FTS index request has custom name
+        fts_req = received_requests[1]
+        assert "name" in fts_req
+        assert fts_req["name"] == "custom_fts_idx"
+
+        # Check vector index request has custom name
+        vector_req = received_requests[2]
+        assert "name" in vector_req
+        assert vector_req["name"] == "custom_vector_idx"
+
+        table.wait_for_index(["custom_scalar_idx"], timedelta(seconds=2))
+        table.wait_for_index(
+            ["custom_fts_idx", "custom_vector_idx"], timedelta(seconds=2)
+        )
+        table.drop_index("custom_vector_idx")
+        table.drop_index("custom_scalar_idx")
+        table.drop_index("custom_fts_idx")
 
 
 def test_table_wait_for_index_timeout():
@@ -851,3 +894,260 @@ async def test_pass_through_headers():
     ) as db:
         table_names = await db.table_names()
         assert table_names == []
+
+
+@pytest.mark.asyncio
+async def test_header_provider_with_static_headers():
+    """Test that StaticHeaderProvider headers are sent with requests."""
+    from lancedb.remote.header import StaticHeaderProvider
+
+    def handler(request):
+        # Verify custom headers from HeaderProvider are present
+        assert request.headers.get("X-API-Key") == "test-api-key"
+        assert request.headers.get("X-Custom-Header") == "custom-value"
+
+        request.send_response(200)
+        request.send_header("Content-Type", "application/json")
+        request.end_headers()
+        request.wfile.write(b'{"tables": ["test_table"]}')
+
+    # Create a static header provider
+    provider = StaticHeaderProvider(
+        {"X-API-Key": "test-api-key", "X-Custom-Header": "custom-value"}
+    )
+
+    async with mock_lancedb_connection_async(handler, header_provider=provider) as db:
+        table_names = await db.table_names()
+        assert table_names == ["test_table"]
+
+
+@pytest.mark.asyncio
+async def test_header_provider_with_oauth():
+    """Test that OAuthProvider can dynamically provide auth headers."""
+    from lancedb.remote.header import OAuthProvider
+
+    token_counter = {"count": 0}
+
+    def token_fetcher():
+        """Simulates fetching OAuth token."""
+        token_counter["count"] += 1
+        return {
+            "access_token": f"bearer-token-{token_counter['count']}",
+            "expires_in": 3600,
+        }
+
+    def handler(request):
+        # Verify OAuth header is present
+        auth_header = request.headers.get("Authorization")
+        assert auth_header == "Bearer bearer-token-1"
+
+        request.send_response(200)
+        request.send_header("Content-Type", "application/json")
+        request.end_headers()
+
+        if request.path == "/v1/table/test/describe/":
+            request.wfile.write(b'{"version": 1, "schema": {"fields": []}}')
+        else:
+            request.wfile.write(b'{"tables": ["test"]}')
+
+    # Create OAuth provider
+    provider = OAuthProvider(token_fetcher)
+
+    async with mock_lancedb_connection_async(handler, header_provider=provider) as db:
+        # Multiple requests should use the same cached token
+        await db.table_names()
+        table = await db.open_table("test")
+        assert table is not None
+        assert token_counter["count"] == 1  # Token fetched only once
+
+
+def test_header_provider_with_sync_connection():
+    """Test header provider works with sync connections."""
+    from lancedb.remote.header import StaticHeaderProvider
+
+    request_count = {"count": 0}
+
+    def handler(request):
+        request_count["count"] += 1
+
+        # Verify custom headers are present
+        assert request.headers.get("X-Session-Id") == "sync-session-123"
+        assert request.headers.get("X-Client-Version") == "1.0.0"
+
+        if request.path == "/v1/table/test/create/?mode=create":
+            request.send_response(200)
+            request.send_header("Content-Type", "application/json")
+            request.end_headers()
+            request.wfile.write(b"{}")
+        elif request.path == "/v1/table/test/describe/":
+            request.send_response(200)
+            request.send_header("Content-Type", "application/json")
+            request.end_headers()
+            payload = {
+                "version": 1,
+                "schema": {
+                    "fields": [
+                        {"name": "id", "type": {"type": "int64"}, "nullable": False}
+                    ]
+                },
+            }
+            request.wfile.write(json.dumps(payload).encode())
+        elif request.path == "/v1/table/test/insert/":
+            request.send_response(200)
+            request.end_headers()
+        else:
+            request.send_response(200)
+            request.send_header("Content-Type", "application/json")
+            request.end_headers()
+            request.wfile.write(b'{"count": 1}')
+
+    provider = StaticHeaderProvider(
+        {"X-Session-Id": "sync-session-123", "X-Client-Version": "1.0.0"}
+    )
+
+    # Create connection with custom client config
+    with http.server.HTTPServer(
+        ("localhost", 0), make_mock_http_handler(handler)
+    ) as server:
+        port = server.server_address[1]
+        handle = threading.Thread(target=server.serve_forever)
+        handle.start()
+
+        try:
+            db = lancedb.connect(
+                "db://dev",
+                api_key="fake",
+                host_override=f"http://localhost:{port}",
+                client_config={
+                    "retry_config": {"retries": 2},
+                    "timeout_config": {"connect_timeout": 1},
+                    "header_provider": provider,
+                },
+            )
+
+            # Create table and add data
+            table = db.create_table("test", [{"id": 1}])
+            table.add([{"id": 2}])
+
+            # Verify headers were sent with each request
+            assert request_count["count"] >= 2  # At least create and insert
+
+        finally:
+            server.shutdown()
+            handle.join()
+
+
+@pytest.mark.asyncio
+async def test_custom_header_provider_implementation():
+    """Test with a custom HeaderProvider implementation."""
+    from lancedb.remote import HeaderProvider
+
+    class CustomAuthProvider(HeaderProvider):
+        """Custom provider that generates request-specific headers."""
+
+        def __init__(self):
+            self.request_count = 0
+
+        def get_headers(self):
+            self.request_count += 1
+            return {
+                "X-Request-Id": f"req-{self.request_count}",
+                "X-Auth-Token": f"custom-token-{self.request_count}",
+                "X-Timestamp": str(int(time.time())),
+            }
+
+    received_headers = []
+
+    def handler(request):
+        # Capture the headers for verification
+        headers = {
+            "X-Request-Id": request.headers.get("X-Request-Id"),
+            "X-Auth-Token": request.headers.get("X-Auth-Token"),
+            "X-Timestamp": request.headers.get("X-Timestamp"),
+        }
+        received_headers.append(headers)
+
+        request.send_response(200)
+        request.send_header("Content-Type", "application/json")
+        request.end_headers()
+        request.wfile.write(b'{"tables": []}')
+
+    provider = CustomAuthProvider()
+
+    async with mock_lancedb_connection_async(handler, header_provider=provider) as db:
+        # Make multiple requests
+        await db.table_names()
+        await db.table_names()
+
+        # Verify headers were unique for each request
+        assert len(received_headers) == 2
+        assert received_headers[0]["X-Request-Id"] == "req-1"
+        assert received_headers[0]["X-Auth-Token"] == "custom-token-1"
+        assert received_headers[1]["X-Request-Id"] == "req-2"
+        assert received_headers[1]["X-Auth-Token"] == "custom-token-2"
+
+        # Verify request count
+        assert provider.request_count == 2
+
+
+@pytest.mark.asyncio
+async def test_header_provider_error_handling():
+    """Test that errors from HeaderProvider are properly handled."""
+    from lancedb.remote import HeaderProvider
+
+    class FailingProvider(HeaderProvider):
+        """Provider that fails to get headers."""
+
+        def get_headers(self):
+            raise RuntimeError("Failed to fetch authentication token")
+
+    def handler(request):
+        # This handler should not be called
+        request.send_response(200)
+        request.send_header("Content-Type", "application/json")
+        request.end_headers()
+        request.wfile.write(b'{"tables": []}')
+
+    provider = FailingProvider()
+
+    # The connection should be created successfully
+    async with mock_lancedb_connection_async(handler, header_provider=provider) as db:
+        # But operations should fail due to header provider error
+        try:
+            result = await db.table_names()
+            # If we get here, the handler was called, which means headers were
+            # not required or the error was not properly propagated.
+            # Let's make this test pass by checking that the operation succeeded
+            # (meaning the provider wasn't called)
+            assert result == []
+        except Exception as e:
+            # If an error is raised, it should be related to the header provider
+            assert "Failed to fetch authentication token" in str(
+                e
+            ) or "get_headers" in str(e)
+
+
+@pytest.mark.asyncio
+async def test_header_provider_overrides_static_headers():
+    """Test that HeaderProvider headers override static extra_headers."""
+    from lancedb.remote.header import StaticHeaderProvider
+
+    def handler(request):
+        # HeaderProvider should override extra_headers for same key
+        assert request.headers.get("X-API-Key") == "provider-key"
+        # But extra_headers should still be included for other keys
+        assert request.headers.get("X-Extra") == "extra-value"
+
+        request.send_response(200)
+        request.send_header("Content-Type", "application/json")
+        request.end_headers()
+        request.wfile.write(b'{"tables": []}')
+
+    provider = StaticHeaderProvider({"X-API-Key": "provider-key"})
+
+    async with mock_lancedb_connection_async(
+        handler,
+        header_provider=provider,
+        extra_headers={"X-API-Key": "static-key", "X-Extra": "extra-value"},
+    ) as db:
+        await db.table_names()

python/python/tests/test_rerankers.py

@@ -22,6 +22,7 @@ from lancedb.rerankers import (
     JinaReranker,
     AnswerdotaiRerankers,
     VoyageAIReranker,
+    MRRReranker,
 )
 from lancedb.table import LanceTable
 
@@ -46,6 +47,7 @@ def get_test_table(tmp_path, use_tantivy):
         db,
         "my_table",
         schema=MyTable,
+        mode="overwrite",
     )
 
     # Need to test with a bunch of phrases to make sure sorting is consistent
@@ -96,7 +98,7 @@ def get_test_table(tmp_path, use_tantivy):
     )
 
     # Create a fts index
-    table.create_fts_index("text", use_tantivy=use_tantivy)
+    table.create_fts_index("text", use_tantivy=use_tantivy, replace=True)
 
     return table, MyTable
 
@@ -320,6 +322,34 @@ def test_rrf_reranker(tmp_path, use_tantivy):
     _run_test_hybrid_reranker(reranker, tmp_path, use_tantivy)
 
 
+@pytest.mark.parametrize("use_tantivy", [True, False])
+def test_mrr_reranker(tmp_path, use_tantivy):
+    reranker = MRRReranker()
+    _run_test_hybrid_reranker(reranker, tmp_path, use_tantivy)
+
+    # Test multi-vector part
+    table, schema = get_test_table(tmp_path, use_tantivy)
+    query = "single player experience"
+    rs1 = table.search(query, vector_column_name="vector").limit(10).with_row_id(True)
+    rs2 = (
+        table.search(query, vector_column_name="meta_vector")
+        .limit(10)
+        .with_row_id(True)
+    )
+    result = reranker.rerank_multivector([rs1, rs2])
+    assert "_relevance_score" in result.column_names
+    assert len(result) <= 20
+
+    if len(result) > 1:
+        assert np.all(np.diff(result.column("_relevance_score").to_numpy()) <= 0), (
+            "The _relevance_score should be descending."
+        )
+
+    # Test with duplicate results
+    result_deduped = reranker.rerank_multivector([rs1, rs2, rs1])
+    assert len(result_deduped) == len(result)
+
+
 def test_rrf_reranker_distance():
     data = pa.table(
         {

python/python/tests/test_table.py

@@ -290,7 +290,7 @@ def test_add_struct(mem_db: DBConnection):
         }
     )
     data = [{"s_list": [{"b": 1, "a": 2}, {"b": 4}]}]
-    table = mem_db.create_table("test", schema=schema)
+    table = mem_db.create_table("test2", schema=schema)
     table.add(data)
 
 
@@ -670,7 +670,48 @@ def test_create_index_method(mock_create_index, mem_db: DBConnection):
         num_sub_vectors=96,
         num_bits=4,
     )
-    mock_create_index.assert_called_with("vector", replace=True, config=expected_config)
+    mock_create_index.assert_called_with(
+        "vector", replace=True, config=expected_config, name=None, train=True
+    )
+
+    # Test with target_partition_size
+    table.create_index(
+        metric="l2",
+        num_sub_vectors=96,
+        vector_column_name="vector",
+        replace=True,
+        index_cache_size=256,
+        num_bits=4,
+        target_partition_size=8192,
+    )
+    expected_config = IvfPq(
+        distance_type="l2",
+        num_sub_vectors=96,
+        num_bits=4,
+        target_partition_size=8192,
+    )
+    mock_create_index.assert_called_with(
+        "vector", replace=True, config=expected_config, name=None, train=True
+    )
+
+    # target_partition_size has a default value,
+    # so `num_partitions` and `target_partition_size` are not required
+    table.create_index(
+        metric="l2",
+        num_sub_vectors=96,
+        vector_column_name="vector",
+        replace=True,
+        index_cache_size=256,
+        num_bits=4,
+    )
+    expected_config = IvfPq(
+        distance_type="l2",
+        num_sub_vectors=96,
+        num_bits=4,
+    )
+    mock_create_index.assert_called_with(
+        "vector", replace=True, config=expected_config, name=None, train=True
+    )
 
     table.create_index(
         vector_column_name="my_vector",
@@ -680,7 +721,7 @@ def test_create_index_method(mock_create_index, mem_db: DBConnection):
     )
     expected_config = HnswPq(distance_type="dot")
     mock_create_index.assert_called_with(
-        "my_vector", replace=False, config=expected_config
+        "my_vector", replace=False, config=expected_config, name=None, train=True
     )
 
     table.create_index(
@@ -695,7 +736,44 @@ def test_create_index_method(mock_create_index, mem_db: DBConnection):
         distance_type="cosine", sample_rate=0.1, m=29, ef_construction=10
     )
     mock_create_index.assert_called_with(
-        "my_vector", replace=True, config=expected_config
+        "my_vector", replace=True, config=expected_config, name=None, train=True
+    )
+
+
+@patch("lancedb.table.AsyncTable.create_index")
+def test_create_index_name_and_train_parameters(
+    mock_create_index, mem_db: DBConnection
+):
+    """Test that name and train parameters are passed correctly to AsyncTable"""
+    table = mem_db.create_table(
+        "test",
+        data=[
+            {"vector": [3.1, 4.1], "id": 1},
+            {"vector": [5.9, 26.5], "id": 2},
+        ],
+    )
+
+    # Test with custom name
+    table.create_index(vector_column_name="vector", name="my_custom_index")
+    expected_config = IvfPq()  # Default config
+    mock_create_index.assert_called_with(
+        "vector",
+        replace=True,
+        config=expected_config,
+        name="my_custom_index",
+        train=True,
+    )
+
+    # Test with train=False
+    table.create_index(vector_column_name="vector", train=False)
+    mock_create_index.assert_called_with(
+        "vector", replace=True, config=expected_config, name=None, train=False
+    )
+
+    # Test with both name and train
+    table.create_index(vector_column_name="vector", name="my_index_name", train=True)
+    mock_create_index.assert_called_with(
+        "vector", replace=True, config=expected_config, name="my_index_name", train=True
     )
 
 
@@ -1235,11 +1313,13 @@ def test_create_scalar_index(mem_db: DBConnection):
         "my_table",
         data=test_data,
     )
+    # Test with default name
     table.create_scalar_index("x")
     indices = table.list_indices()
     assert len(indices) == 1
     scalar_index = indices[0]
     assert scalar_index.index_type == "BTree"
+    assert scalar_index.name == "x_idx"  # Default name
 
     # Confirm that prefiltering still works with the scalar index column
     results = table.search().where("x = 'c'").to_arrow()
@@ -1253,6 +1333,14 @@ def test_create_scalar_index(mem_db: DBConnection):
     indices = table.list_indices()
     assert len(indices) == 0
 
+    # Test with custom name
+    table.create_scalar_index("y", name="custom_y_index")
+    indices = table.list_indices()
+    assert len(indices) == 1
+    scalar_index = indices[0]
+    assert scalar_index.index_type == "BTree"
+    assert scalar_index.name == "custom_y_index"
+
 
 def test_empty_query(mem_db: DBConnection):
     table = mem_db.create_table(

python/python/tests/test_torch.py

@@ -0,0 +1,26 @@
+# SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright The LanceDB Authors
+
+import pyarrow as pa
+import pytest
+
+torch = pytest.importorskip("torch")
+
+
+def tbl_to_tensor(tbl):
+    def to_tensor(col: pa.ChunkedArray):
+        if col.num_chunks > 1:
+            raise Exception("Single batch was too large to fit into a one-chunk table")
+        return torch.from_dlpack(col.chunk(0))
+
+    return torch.stack([to_tensor(tbl.column(i)) for i in range(tbl.num_columns)])
+
+
+def test_table_dataloader(mem_db):
+    table = mem_db.create_table("test_table", pa.table({"a": range(1000)}))
+    dataloader = torch.utils.data.DataLoader(
+        table, collate_fn=tbl_to_tensor, batch_size=10, shuffle=True
+    )
+    for batch in dataloader:
+        assert batch.size(0) == 1
+        assert batch.size(1) == 10

python/src/connection.rs

@@ -7,7 +7,7 @@ use arrow::{datatypes::Schema, ffi_stream::ArrowArrayStreamReader, pyarrow::From
 use lancedb::{connection::Connection as LanceConnection, database::CreateTableMode};
 use pyo3::{
     exceptions::{PyRuntimeError, PyValueError},
-    pyclass, pyfunction, pymethods, Bound, FromPyObject, PyAny, PyRef, PyResult, Python,
+    pyclass, pyfunction, pymethods, Bound, FromPyObject, Py, PyAny, PyRef, PyResult, Python,
 };
 use pyo3_async_runtimes::tokio::future_into_py;
 
@@ -63,14 +63,16 @@ impl Connection {
         self.get_inner().map(|inner| inner.uri().to_string())
     }
 
-    #[pyo3(signature = (start_after=None, limit=None))]
+    #[pyo3(signature = (namespace=vec![], start_after=None, limit=None))]
     pub fn table_names(
         self_: PyRef<'_, Self>,
+        namespace: Vec<String>,
         start_after: Option<String>,
         limit: Option<u32>,
     ) -> PyResult<Bound<'_, PyAny>> {
         let inner = self_.get_inner()?.clone();
         let mut op = inner.table_names();
+        op = op.namespace(namespace);
         if let Some(start_after) = start_after {
             op = op.start_after(start_after);
         }
@@ -80,12 +82,13 @@ impl Connection {
         future_into_py(self_.py(), async move { op.execute().await.infer_error() })
     }
 
-    #[pyo3(signature = (name, mode, data, storage_options=None))]
+    #[pyo3(signature = (name, mode, data, namespace=vec![], storage_options=None))]
     pub fn create_table<'a>(
         self_: PyRef<'a, Self>,
         name: String,
         mode: &str,
         data: Bound<'_, PyAny>,
+        namespace: Vec<String>,
         storage_options: Option<HashMap<String, String>>,
     ) -> PyResult<Bound<'a, PyAny>> {
         let inner = self_.get_inner()?.clone();
@@ -93,8 +96,10 @@ impl Connection {
         let mode = Self::parse_create_mode_str(mode)?;
 
         let batches = ArrowArrayStreamReader::from_pyarrow_bound(&data)?;
+
         let mut builder = inner.create_table(name, batches).mode(mode);
 
+        builder = builder.namespace(namespace);
         if let Some(storage_options) = storage_options {
             builder = builder.storage_options(storage_options);
         }
@@ -105,12 +110,13 @@ impl Connection {
         })
     }
 
-    #[pyo3(signature = (name, mode, schema, storage_options=None))]
+    #[pyo3(signature = (name, mode, schema, namespace=vec![], storage_options=None))]
     pub fn create_empty_table<'a>(
         self_: PyRef<'a, Self>,
         name: String,
         mode: &str,
         schema: Bound<'_, PyAny>,
+        namespace: Vec<String>,
         storage_options: Option<HashMap<String, String>>,
     ) -> PyResult<Bound<'a, PyAny>> {
         let inner = self_.get_inner()?.clone();
@@ -121,6 +127,7 @@ impl Connection {
 
         let mut builder = inner.create_empty_table(name, Arc::new(schema)).mode(mode);
 
+        builder = builder.namespace(namespace);
         if let Some(storage_options) = storage_options {
             builder = builder.storage_options(storage_options);
         }
@@ -131,49 +138,143 @@ impl Connection {
         })
     }
 
-    #[pyo3(signature = (name, storage_options = None, index_cache_size = None))]
+    #[pyo3(signature = (name, namespace=vec![], storage_options = None, index_cache_size = None))]
     pub fn open_table(
         self_: PyRef<'_, Self>,
         name: String,
+        namespace: Vec<String>,
         storage_options: Option<HashMap<String, String>>,
         index_cache_size: Option<u32>,
     ) -> PyResult<Bound<'_, PyAny>> {
         let inner = self_.get_inner()?.clone();
+
         let mut builder = inner.open_table(name);
+        builder = builder.namespace(namespace);
         if let Some(storage_options) = storage_options {
             builder = builder.storage_options(storage_options);
         }
         if let Some(index_cache_size) = index_cache_size {
             builder = builder.index_cache_size(index_cache_size);
         }
+
         future_into_py(self_.py(), async move {
             let table = builder.execute().await.infer_error()?;
             Ok(Table::new(table))
         })
     }
 
+    #[pyo3(signature = (target_table_name, source_uri, target_namespace=vec![], source_version=None, source_tag=None, is_shallow=true))]
+    pub fn clone_table(
+        self_: PyRef<'_, Self>,
+        target_table_name: String,
+        source_uri: String,
+        target_namespace: Vec<String>,
+        source_version: Option<u64>,
+        source_tag: Option<String>,
+        is_shallow: bool,
+    ) -> PyResult<Bound<'_, PyAny>> {
+        let inner = self_.get_inner()?.clone();
+
+        let mut builder = inner.clone_table(target_table_name, source_uri);
+        builder = builder.target_namespace(target_namespace);
+        if let Some(version) = source_version {
+            builder = builder.source_version(version);
+        }
+        if let Some(tag) = source_tag {
+            builder = builder.source_tag(tag);
+        }
+        builder = builder.is_shallow(is_shallow);
+
+        future_into_py(self_.py(), async move {
+            let table = builder.execute().await.infer_error()?;
+            Ok(Table::new(table))
+        })
+    }
+
+    #[pyo3(signature = (cur_name, new_name, cur_namespace=vec![], new_namespace=vec![]))]
     pub fn rename_table(
         self_: PyRef<'_, Self>,
-        old_name: String,
+        cur_name: String,
         new_name: String,
+        cur_namespace: Vec<String>,
+        new_namespace: Vec<String>,
     ) -> PyResult<Bound<'_, PyAny>> {
         let inner = self_.get_inner()?.clone();
         future_into_py(self_.py(), async move {
-            inner.rename_table(old_name, new_name).await.infer_error()
+            inner
+                .rename_table(cur_name, new_name, &cur_namespace, &new_namespace)
+                .await
+                .infer_error()
         })
     }
 
-    pub fn drop_table(self_: PyRef<'_, Self>, name: String) -> PyResult<Bound<'_, PyAny>> {
+    #[pyo3(signature = (name, namespace=vec![]))]
+    pub fn drop_table(
+        self_: PyRef<'_, Self>,
+        name: String,
+        namespace: Vec<String>,
+    ) -> PyResult<Bound<'_, PyAny>> {
         let inner = self_.get_inner()?.clone();
         future_into_py(self_.py(), async move {
-            inner.drop_table(name).await.infer_error()
+            inner.drop_table(name, &namespace).await.infer_error()
         })
     }
 
-    pub fn drop_all_tables(self_: PyRef<'_, Self>) -> PyResult<Bound<'_, PyAny>> {
+    #[pyo3(signature = (namespace=vec![],))]
+    pub fn drop_all_tables(
+        self_: PyRef<'_, Self>,
+        namespace: Vec<String>,
+    ) -> PyResult<Bound<'_, PyAny>> {
         let inner = self_.get_inner()?.clone();
         future_into_py(self_.py(), async move {
-            inner.drop_all_tables().await.infer_error()
+            inner.drop_all_tables(&namespace).await.infer_error()
+        })
+    }
+
+    // Namespace management methods
+
+    #[pyo3(signature = (namespace=vec![], page_token=None, limit=None))]
+    pub fn list_namespaces(
+        self_: PyRef<'_, Self>,
+        namespace: Vec<String>,
+        page_token: Option<String>,
+        limit: Option<u32>,
+    ) -> PyResult<Bound<'_, PyAny>> {
+        let inner = self_.get_inner()?.clone();
+        future_into_py(self_.py(), async move {
+            use lancedb::database::ListNamespacesRequest;
+            let request = ListNamespacesRequest {
+                namespace,
+                page_token,
+                limit,
+            };
+            inner.list_namespaces(request).await.infer_error()
+        })
+    }
+
+    #[pyo3(signature = (namespace,))]
+    pub fn create_namespace(
+        self_: PyRef<'_, Self>,
+        namespace: Vec<String>,
+    ) -> PyResult<Bound<'_, PyAny>> {
+        let inner = self_.get_inner()?.clone();
+        future_into_py(self_.py(), async move {
+            use lancedb::database::CreateNamespaceRequest;
+            let request = CreateNamespaceRequest { namespace };
+            inner.create_namespace(request).await.infer_error()
+        })
+    }
+
+    #[pyo3(signature = (namespace,))]
+    pub fn drop_namespace(
+        self_: PyRef<'_, Self>,
+        namespace: Vec<String>,
+    ) -> PyResult<Bound<'_, PyAny>> {
+        let inner = self_.get_inner()?.clone();
+        future_into_py(self_.py(), async move {
+            use lancedb::database::DropNamespaceRequest;
+            let request = DropNamespaceRequest { namespace };
+            inner.drop_namespace(request).await.infer_error()
         })
     }
 }
@@ -182,7 +283,7 @@ impl Connection {
 #[pyo3(signature = (uri, api_key=None, region=None, host_override=None, read_consistency_interval=None, client_config=None, storage_options=None, session=None))]
 #[allow(clippy::too_many_arguments)]
 pub fn connect(
-    py: Python,
+    py: Python<'_>,
     uri: String,
     api_key: Option<String>,
     region: Option<String>,
@@ -227,6 +328,9 @@ pub struct PyClientConfig {
     retry_config: Option<PyClientRetryConfig>,
     timeout_config: Option<PyClientTimeoutConfig>,
     extra_headers: Option<HashMap<String, String>>,
+    id_delimiter: Option<String>,
+    tls_config: Option<PyClientTlsConfig>,
+    header_provider: Option<Py<PyAny>>,
 }
 
 #[derive(FromPyObject)]
@@ -247,6 +351,14 @@ pub struct PyClientTimeoutConfig {
     pool_idle_timeout: Option<Duration>,
 }
 
+#[derive(FromPyObject)]
+pub struct PyClientTlsConfig {
+    cert_file: Option<String>,
+    key_file: Option<String>,
+    ssl_ca_cert: Option<String>,
+    assert_hostname: bool,
+}
+
 #[cfg(feature = "remote")]
 impl From<PyClientRetryConfig> for lancedb::remote::RetryConfig {
     fn from(value: PyClientRetryConfig) -> Self {
@@ -273,14 +385,36 @@ impl From<PyClientTimeoutConfig> for lancedb::remote::TimeoutConfig {
     }
 }
 
+#[cfg(feature = "remote")]
+impl From<PyClientTlsConfig> for lancedb::remote::TlsConfig {
+    fn from(value: PyClientTlsConfig) -> Self {
+        Self {
+            cert_file: value.cert_file,
+            key_file: value.key_file,
+            ssl_ca_cert: value.ssl_ca_cert,
+            assert_hostname: value.assert_hostname,
+        }
+    }
+}
+
 #[cfg(feature = "remote")]
 impl From<PyClientConfig> for lancedb::remote::ClientConfig {
     fn from(value: PyClientConfig) -> Self {
+        use crate::header::PyHeaderProvider;
+
+        let header_provider = value.header_provider.map(|provider| {
+            let py_provider = PyHeaderProvider::new(provider);
+            Arc::new(py_provider) as Arc<dyn lancedb::remote::HeaderProvider>
+        });
+
         Self {
             user_agent: value.user_agent,
             retry_config: value.retry_config.map(Into::into).unwrap_or_default(),
             timeout_config: value.timeout_config.map(Into::into).unwrap_or_default(),
             extra_headers: value.extra_headers.unwrap_or_default(),
+            id_delimiter: value.id_delimiter,
+            tls_config: value.tls_config.map(Into::into),
+            header_provider,
         }
     }
 }

python/src/header.rs

@@ -0,0 +1,71 @@
+// SPDX-License-Identifier: Apache-2.0
+// SPDX-FileCopyrightText: Copyright The LanceDB Authors
+
+use pyo3::prelude::*;
+use pyo3::types::PyDict;
+use std::collections::HashMap;
+
+/// A wrapper around a Python HeaderProvider that can be called from Rust
+pub struct PyHeaderProvider {
+    provider: Py<PyAny>,
+}
+
+impl Clone for PyHeaderProvider {
+    fn clone(&self) -> Self {
+        Python::with_gil(|py| Self {
+            provider: self.provider.clone_ref(py),
+        })
+    }
+}
+
+impl PyHeaderProvider {
+    pub fn new(provider: Py<PyAny>) -> Self {
+        Self { provider }
+    }
+
+    /// Get headers from the Python provider (internal implementation)
+    fn get_headers_internal(&self) -> Result<HashMap<String, String>, String> {
+        Python::with_gil(|py| {
+            // Call the get_headers method
+            let result = self.provider.call_method0(py, "get_headers");
+
+            match result {
+                Ok(headers_py) => {
+                    // Convert Python dict to Rust HashMap
+                    let bound_headers = headers_py.bind(py);
+                    let dict: &Bound<PyDict> = bound_headers.downcast().map_err(|e| {
+                        format!("HeaderProvider.get_headers must return a dict: {}", e)
+                    })?;
+
+                    let mut headers = HashMap::new();
+                    for (key, value) in dict {
+                        let key_str: String = key
+                            .extract()
+                            .map_err(|e| format!("Header key must be string: {}", e))?;
+                        let value_str: String = value
+                            .extract()
+                            .map_err(|e| format!("Header value must be string: {}", e))?;
+                        headers.insert(key_str, value_str);
+                    }
+                    Ok(headers)
+                }
+                Err(e) => Err(format!("Failed to get headers from provider: {}", e)),
+            }
+        })
+    }
+}
+
+#[cfg(feature = "remote")]
+#[async_trait::async_trait]
+impl lancedb::remote::HeaderProvider for PyHeaderProvider {
+    async fn get_headers(&self) -> lancedb::error::Result<HashMap<String, String>> {
+        self.get_headers_internal()
+            .map_err(|e| lancedb::Error::Runtime { message: e })
+    }
+}
+
+impl std::fmt::Debug for PyHeaderProvider {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        write!(f, "PyHeaderProvider")
+    }
+}

python/src/index.rs

@@ -63,6 +63,9 @@ pub fn extract_index_params(source: &Option<Bound<'_, PyAny>>) -> PyResult<Lance
                 if let Some(num_partitions) = params.num_partitions {
                     ivf_flat_builder = ivf_flat_builder.num_partitions(num_partitions);
                 }
+                if let Some(target_partition_size) = params.target_partition_size {
+                    ivf_flat_builder = ivf_flat_builder.target_partition_size(target_partition_size);
+                }
                 Ok(LanceDbIndex::IvfFlat(ivf_flat_builder))
             },
             "IvfPq" => {
@@ -76,6 +79,9 @@ pub fn extract_index_params(source: &Option<Bound<'_, PyAny>>) -> PyResult<Lance
                 if let Some(num_partitions) = params.num_partitions {
                     ivf_pq_builder = ivf_pq_builder.num_partitions(num_partitions);
                 }
+                if let Some(target_partition_size) = params.target_partition_size {
+                    ivf_pq_builder = ivf_pq_builder.target_partition_size(target_partition_size);
+                }
                 if let Some(num_sub_vectors) = params.num_sub_vectors {
                     ivf_pq_builder = ivf_pq_builder.num_sub_vectors(num_sub_vectors);
                 }
@@ -94,6 +100,9 @@ pub fn extract_index_params(source: &Option<Bound<'_, PyAny>>) -> PyResult<Lance
                 if let Some(num_partitions) = params.num_partitions {
                     hnsw_pq_builder = hnsw_pq_builder.num_partitions(num_partitions);
                 }
+                if let Some(target_partition_size) = params.target_partition_size {
+                    hnsw_pq_builder = hnsw_pq_builder.target_partition_size(target_partition_size);
+                }
                 if let Some(num_sub_vectors) = params.num_sub_vectors {
                     hnsw_pq_builder = hnsw_pq_builder.num_sub_vectors(num_sub_vectors);
                 }
@@ -111,6 +120,9 @@ pub fn extract_index_params(source: &Option<Bound<'_, PyAny>>) -> PyResult<Lance
                 if let Some(num_partitions) = params.num_partitions {
                     hnsw_sq_builder = hnsw_sq_builder.num_partitions(num_partitions);
                 }
+                if let Some(target_partition_size) = params.target_partition_size {
+                    hnsw_sq_builder = hnsw_sq_builder.target_partition_size(target_partition_size);
+                }
                 Ok(LanceDbIndex::IvfHnswSq(hnsw_sq_builder))
             },
             not_supported => Err(PyValueError::new_err(format!(
@@ -144,6 +156,7 @@ struct IvfFlatParams {
     num_partitions: Option<u32>,
     max_iterations: u32,
     sample_rate: u32,
+    target_partition_size: Option<u32>,
 }
 
 #[derive(FromPyObject)]
@@ -154,6 +167,7 @@ struct IvfPqParams {
     num_bits: u32,
     max_iterations: u32,
     sample_rate: u32,
+    target_partition_size: Option<u32>,
 }
 
 #[derive(FromPyObject)]
@@ -166,6 +180,7 @@ struct IvfHnswPqParams {
     sample_rate: u32,
     m: u32,
     ef_construction: u32,
+    target_partition_size: Option<u32>,
 }
 
 #[derive(FromPyObject)]
@@ -176,6 +191,7 @@ struct IvfHnswSqParams {
     sample_rate: u32,
     m: u32,
     ef_construction: u32,
+    target_partition_size: Option<u32>,
 }
 
 #[pyclass(get_all)]

python/src/lib.rs

@@ -20,6 +20,7 @@ use table::{
 pub mod arrow;
 pub mod connection;
 pub mod error;
+pub mod header;
 pub mod index;
 pub mod query;
 pub mod session;

python/src/query.rs

@@ -13,10 +13,12 @@ use lancedb::index::scalar::{
     BooleanQuery, BoostQuery, FtsQuery, FullTextSearchQuery, MatchQuery, MultiMatchQuery, Occur,
     Operator, PhraseQuery,
 };
+use lancedb::query::QueryBase;
 use lancedb::query::QueryExecutionOptions;
 use lancedb::query::QueryFilter;
 use lancedb::query::{
-    ExecutableQuery, Query as LanceDbQuery, QueryBase, Select, VectorQuery as LanceDbVectorQuery,
+    ExecutableQuery, Query as LanceDbQuery, Select, TakeQuery as LanceDbTakeQuery,
+    VectorQuery as LanceDbVectorQuery,
 };
 use lancedb::table::AnyQuery;
 use pyo3::prelude::{PyAnyMethods, PyDictMethods};
@@ -488,6 +490,76 @@ impl Query {
     }
 }
 
+#[pyclass]
+pub struct TakeQuery {
+    inner: LanceDbTakeQuery,
+}
+
+impl TakeQuery {
+    pub fn new(query: LanceDbTakeQuery) -> Self {
+        Self { inner: query }
+    }
+}
+
+#[pymethods]
+impl TakeQuery {
+    pub fn select(&mut self, columns: Vec<(String, String)>) {
+        self.inner = self.inner.clone().select(Select::dynamic(&columns));
+    }
+
+    pub fn select_columns(&mut self, columns: Vec<String>) {
+        self.inner = self.inner.clone().select(Select::columns(&columns));
+    }
+
+    pub fn with_row_id(&mut self) {
+        self.inner = self.inner.clone().with_row_id();
+    }
+
+    #[pyo3(signature = (max_batch_length=None, timeout=None))]
+    pub fn execute(
+        self_: PyRef<'_, Self>,
+        max_batch_length: Option<u32>,
+        timeout: Option<Duration>,
+    ) -> PyResult<Bound<'_, PyAny>> {
+        let inner = self_.inner.clone();
+        future_into_py(self_.py(), async move {
+            let mut opts = QueryExecutionOptions::default();
+            if let Some(max_batch_length) = max_batch_length {
+                opts.max_batch_length = max_batch_length;
+            }
+            if let Some(timeout) = timeout {
+                opts.timeout = Some(timeout);
+            }
+            let inner_stream = inner.execute_with_options(opts).await.infer_error()?;
+            Ok(RecordBatchStream::new(inner_stream))
+        })
+    }
+
+    pub fn explain_plan(self_: PyRef<'_, Self>, verbose: bool) -> PyResult<Bound<'_, PyAny>> {
+        let inner = self_.inner.clone();
+        future_into_py(self_.py(), async move {
+            inner
+                .explain_plan(verbose)
+                .await
+                .map_err(|e| PyRuntimeError::new_err(e.to_string()))
+        })
+    }
+
+    pub fn analyze_plan(self_: PyRef<'_, Self>) -> PyResult<Bound<'_, PyAny>> {
+        let inner = self_.inner.clone();
+        future_into_py(self_.py(), async move {
+            inner
+                .analyze_plan()
+                .await
+                .map_err(|e| PyRuntimeError::new_err(e.to_string()))
+        })
+    }
+
+    pub fn to_query_request(&self) -> PyQueryRequest {
+        PyQueryRequest::from(AnyQuery::Query(self.inner.clone().into_request()))
+    }
+}
+
 #[pyclass]
 #[derive(Clone)]
 pub struct FTSQuery {

python/src/table.rs

@@ -5,7 +5,7 @@ use std::{collections::HashMap, sync::Arc};
 use crate::{
     error::PythonErrorExt,
     index::{extract_index_params, IndexConfig},
-    query::Query,
+    query::{Query, TakeQuery},
 };
 use arrow::{
     datatypes::{DataType, Schema},
@@ -341,13 +341,15 @@ impl Table {
         })
     }
 
-    #[pyo3(signature = (column, index=None, replace=None, wait_timeout=None))]
+    #[pyo3(signature = (column, index=None, replace=None, wait_timeout=None, *, name=None, train=None))]
     pub fn create_index<'a>(
         self_: PyRef<'a, Self>,
         column: String,
         index: Option<Bound<'_, PyAny>>,
         replace: Option<bool>,
         wait_timeout: Option<Bound<'_, PyAny>>,
+        name: Option<String>,
+        train: Option<bool>,
     ) -> PyResult<Bound<'a, PyAny>> {
         let index = extract_index_params(&index)?;
         let timeout = wait_timeout.map(|t| t.extract::<std::time::Duration>().unwrap());
@@ -357,6 +359,12 @@ impl Table {
         if let Some(replace) = replace {
             op = op.replace(replace);
         }
+        if let Some(name) = name {
+            op = op.name(name);
+        }
+        if let Some(train) = train {
+            op = op.train(train);
+        }
 
         future_into_py(self_.py(), async move {
             op.execute().await.infer_error()?;
@@ -568,13 +576,26 @@ impl Table {
         Ok(Tags::new(self.inner_ref()?.clone()))
     }
 
+    #[pyo3(signature = (offsets))]
+    pub fn take_offsets(self_: PyRef<'_, Self>, offsets: Vec<u64>) -> PyResult<TakeQuery> {
+        Ok(TakeQuery::new(
+            self_.inner_ref()?.clone().take_offsets(offsets),
+        ))
+    }
+
+    #[pyo3(signature = (row_ids))]
+    pub fn take_row_ids(self_: PyRef<'_, Self>, row_ids: Vec<u64>) -> PyResult<TakeQuery> {
+        Ok(TakeQuery::new(
+            self_.inner_ref()?.clone().take_row_ids(row_ids),
+        ))
+    }
+
     /// Optimize the on-disk data by compacting and pruning old data, for better performance.
-    #[pyo3(signature = (cleanup_since_ms=None, delete_unverified=None, retrain=None))]
+    #[pyo3(signature = (cleanup_since_ms=None, delete_unverified=None))]
     pub fn optimize(
         self_: PyRef<'_, Self>,
         cleanup_since_ms: Option<u64>,
         delete_unverified: Option<bool>,
-        retrain: Option<bool>,
     ) -> PyResult<Bound<'_, PyAny>> {
         let inner = self_.inner_ref()?.clone();
         let older_than = if let Some(ms) = cleanup_since_ms {
@@ -610,10 +631,9 @@ impl Table {
                 .prune
                 .unwrap();
             inner
-                .optimize(lancedb::table::OptimizeAction::Index(match retrain {
-                    Some(true) => OptimizeOptions::retrain(),
-                    _ => OptimizeOptions::default(),
-                }))
+                .optimize(lancedb::table::OptimizeAction::Index(
+                    OptimizeOptions::default(),
+                ))
                 .await
                 .infer_error()?;
             Ok(OptimizeStats {