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_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

.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:

CLAUDE.md

@@ -13,10 +13,68 @@ Project layout:
 
 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

@@ -15,14 +15,14 @@ categories = ["database-implementations"]
 rust-version = "1.78.0"
 
 [workspace.dependencies]
-lance = { "version" = "=0.33.0", "features" = ["dynamodb"] }
-lance-io = "=0.33.0"
-lance-index = "=0.33.0"
-lance-linalg = "=0.33.0"
-lance-table = "=0.33.0"
-lance-testing = "=0.33.0"
-lance-datafusion = "=0.33.0"
-lance-encoding = "=0.33.0"
+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"
@@ -33,12 +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",

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)
 

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/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/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

@@ -20,16 +20,20 @@
 - [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)
@@ -74,6 +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/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

@@ -857,6 +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/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,
@@ -93,6 +100,13 @@ export {
   ColumnAlteration,
 } from "./table";
 
+export {
+  HeaderProvider,
+  StaticHeaderProvider,
+  OAuthHeaderProvider,
+  TokenResponse,
+} from "./header";
+
 export { MergeInsertBuilder, WriteExecutionOptions } from "./merge";
 
 export * as embedding from "./embedding";
@@ -131,11 +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.
@@ -169,18 +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) {
@@ -191,6 +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/table.ts

@@ -662,6 +662,8 @@ export class LocalTable extends Table {
       column,
       options?.replace,
       options?.waitTimeoutSeconds,
+      options?.name,
+      options?.train,
     );
   }
 

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

@@ -480,6 +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

@@ -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()
     }
 

python/.bumpversion.toml

@@ -1,5 +1,5 @@
 [tool.bumpversion]
-current_version = "0.24.3"
+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 = "0.24.3"
+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

@@ -943,20 +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
         -------

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

@@ -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,
             )
         )
 

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

@@ -689,6 +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.
 
@@ -721,6 +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
 
@@ -776,6 +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.
 
@@ -790,6 +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
         --------
 
@@ -852,6 +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.
 
@@ -916,6 +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
 
@@ -1106,7 +1120,9 @@ class Table(ABC):
         raise NotImplementedError
 
     @abstractmethod
-    def take_offsets(self, offsets: list[int]) -> LanceTakeQueryBuilder:
+    def take_offsets(
+        self, offsets: list[int], *, with_row_id: bool = False
+    ) -> LanceTakeQueryBuilder:
         """
         Take a list of offsets from the table.
 
@@ -1132,8 +1148,60 @@ class Table(ABC):
             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]) -> LanceTakeQueryBuilder:
+    def take_row_ids(
+        self, row_ids: list[int], *, with_row_id: bool = False
+    ) -> LanceTakeQueryBuilder:
         """
         Take a list of row ids from the table.
 
@@ -1402,10 +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
         ----------------
@@ -1639,13 +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,
             )
@@ -1656,8 +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:
@@ -1929,6 +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:
@@ -1945,6 +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
@@ -1954,6 +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(
@@ -1963,6 +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(
@@ -1974,6 +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(
@@ -1983,6 +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}")
@@ -1992,6 +2069,8 @@ class LanceTable(Table):
                 vector_column_name,
                 replace=replace,
                 config=config,
+                name=name,
+                train=train,
             )
         )
 
@@ -2036,6 +2115,7 @@ class LanceTable(Table):
         *,
         replace: bool = True,
         index_type: ScalarIndexType = "BTREE",
+        name: Optional[str] = None,
     ):
         if index_type == "BTREE":
             config = BTree()
@@ -2046,7 +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(
@@ -2070,6 +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):
@@ -2107,6 +2188,7 @@ class LanceTable(Table):
                     field_names,
                     replace=replace,
                     config=config,
+                    name=name,
                 )
             )
             return
@@ -2473,6 +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,
@@ -2532,6 +2615,7 @@ class LanceTable(Table):
         """
         self = cls.__new__(cls)
         self._conn = db
+        self._namespace = namespace
 
         if data_storage_version is not None:
             warnings.warn(
@@ -2564,6 +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,
             )
         )
@@ -2747,10 +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
         ----------------
@@ -3251,6 +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
 
@@ -3277,6 +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(
@@ -3288,7 +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):
@@ -4198,10 +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
         ----------------
@@ -4224,10 +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
@@ -1355,6 +1356,27 @@ def test_take_queries(tmp_path):
     ]
 
 
+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

@@ -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/table.rs

@@ -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()?;
@@ -583,12 +591,11 @@ impl Table {
     }
 
     /// 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 {
@@ -624,10 +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 {

rust/lancedb/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "lancedb"
-version = "0.21.2"
+version = "0.22.1-beta.3"
 edition.workspace = true
 description = "LanceDB: A serverless, low-latency vector database for AI applications"
 license.workspace = true
@@ -86,18 +86,23 @@ rand = { version = "0.9", features = ["small_rng"] }
 random_word = { version = "0.4.3", features = ["en"] }
 uuid = { version = "1.7.0", features = ["v4"] }
 walkdir = "2"
-aws-sdk-dynamodb = { version = "1.38.0" }
-aws-sdk-s3 = { version = "1.38.0" }
-aws-sdk-kms = { version = "1.37" }
-aws-config = { version = "1.0" }
-aws-smithy-runtime = { version = "1.3" }
+aws-sdk-dynamodb = { version = "1.55.0" }
+aws-sdk-s3 = { version = "1.55.0" }
+aws-sdk-kms = { version = "1.48.0" }
+aws-config = { version = "1.5.10" }
+aws-smithy-runtime = { version = "1.9.1" }
 datafusion.workspace = true
 http-body = "1"                                        # Matching reqwest
 rstest = "0.23.0"
 
 
 [features]
-default = []
+default = ["aws", "gcs", "azure", "dynamodb", "oss"]
+aws = ["lance/aws", "lance-io/aws"]
+oss = ["lance/oss", "lance-io/oss"]
+gcs = ["lance/gcp", "lance-io/gcp"]
+azure = ["lance/azure", "lance-io/azure"]
+dynamodb = ["lance/dynamodb", "aws"]
 remote = ["dep:reqwest", "dep:http", "dep:rand", "dep:uuid"]
 fp16kernels = ["lance-linalg/fp16kernels"]
 s3-test = []

rust/lancedb/examples/hybrid_search.rs

@@ -62,10 +62,8 @@ async fn main() -> Result<()> {
             .as_any()
             .downcast_ref::<StringArray>()
             .unwrap();
-        for text in out.iter() {
-            if let Some(text) = text {
-                println!("Result: {}", text);
-            }
+        for text in out.iter().flatten() {
+            println!("Result: {}", text);
         }
     }
 

rust/lancedb/examples/simple.rs

@@ -43,7 +43,7 @@ async fn main() -> Result<()> {
     // --8<-- [end:delete]
 
     // --8<-- [start:drop_table]
-    db.drop_table("my_table").await.unwrap();
+    db.drop_table("my_table", &[]).await.unwrap();
     // --8<-- [end:drop_table]
     Ok(())
 }

rust/lancedb/src/catalog.rs

@@ -1,86 +0,0 @@
-// SPDX-License-Identifier: Apache-2.0
-// SPDX-FileCopyrightText: Copyright The LanceDB Authors
-
-//! Catalog implementation for managing databases
-
-pub mod listing;
-
-use std::collections::HashMap;
-use std::sync::Arc;
-
-use crate::database::Database;
-use crate::error::Result;
-use async_trait::async_trait;
-
-pub trait CatalogOptions {
-    fn serialize_into_map(&self, map: &mut HashMap<String, String>);
-}
-
-/// Request parameters for listing databases
-#[derive(Clone, Debug, Default)]
-pub struct DatabaseNamesRequest {
-    /// Start listing after this name (exclusive)
-    pub start_after: Option<String>,
-    /// Maximum number of names to return
-    pub limit: Option<u32>,
-}
-
-/// Request to open an existing database
-#[derive(Clone, Debug)]
-pub struct OpenDatabaseRequest {
-    /// The name of the database to open
-    pub name: String,
-    /// A map of database-specific options
-    ///
-    /// Consult the catalog / database implementation to determine which options are available
-    pub database_options: HashMap<String, String>,
-}
-
-/// Database creation mode
-///
-/// The default behavior is Create
-pub enum CreateDatabaseMode {
-    /// Create new database, error if exists
-    Create,
-    /// Open existing database if present
-    ExistOk,
-    /// Overwrite existing database
-    Overwrite,
-}
-
-impl Default for CreateDatabaseMode {
-    fn default() -> Self {
-        Self::Create
-    }
-}
-
-/// Request to create a new database
-pub struct CreateDatabaseRequest {
-    /// The name of the database to create
-    pub name: String,
-    /// The creation mode
-    pub mode: CreateDatabaseMode,
-    /// A map of catalog-specific options, consult your catalog implementation to determine what's available
-    pub options: HashMap<String, String>,
-}
-
-#[async_trait]
-pub trait Catalog: Send + Sync + std::fmt::Debug + 'static {
-    /// List database names with pagination
-    async fn database_names(&self, request: DatabaseNamesRequest) -> Result<Vec<String>>;
-
-    /// Create a new database
-    async fn create_database(&self, request: CreateDatabaseRequest) -> Result<Arc<dyn Database>>;
-
-    /// Open existing database
-    async fn open_database(&self, request: OpenDatabaseRequest) -> Result<Arc<dyn Database>>;
-
-    /// Rename database
-    async fn rename_database(&self, old_name: &str, new_name: &str) -> Result<()>;
-
-    /// Delete database
-    async fn drop_database(&self, name: &str) -> Result<()>;
-
-    /// Delete all databases
-    async fn drop_all_databases(&self) -> Result<()>;
-}

rust/lancedb/src/catalog/listing.rs

@@ -1,622 +0,0 @@
-// SPDX-License-Identifier: Apache-2.0
-// SPDX-FileCopyrightText: Copyright The LanceDB Authors
-
-//! Catalog implementation based on a local file system.
-
-use std::collections::HashMap;
-use std::fs::create_dir_all;
-use std::path::Path;
-use std::sync::Arc;
-
-use super::{
-    Catalog, CatalogOptions, CreateDatabaseMode, CreateDatabaseRequest, DatabaseNamesRequest,
-    OpenDatabaseRequest,
-};
-use crate::connection::ConnectRequest;
-use crate::database::listing::{ListingDatabase, ListingDatabaseOptions};
-use crate::database::{Database, DatabaseOptions};
-use crate::error::{CreateDirSnafu, Error, Result};
-use async_trait::async_trait;
-use lance::io::{ObjectStore, ObjectStoreParams, ObjectStoreRegistry};
-use lance_io::local::to_local_path;
-use object_store::path::Path as ObjectStorePath;
-use snafu::ResultExt;
-
-/// Options for the listing catalog
-///
-/// Note: the catalog will use the `storage_options` configured on
-/// db_options to configure storage for listing / creating / deleting
-/// databases.
-#[derive(Clone, Debug, Default)]
-pub struct ListingCatalogOptions {
-    /// The options to use for databases opened by this catalog
-    ///
-    /// This also contains the storage options used by the catalog
-    pub db_options: ListingDatabaseOptions,
-}
-
-impl CatalogOptions for ListingCatalogOptions {
-    fn serialize_into_map(&self, map: &mut HashMap<String, String>) {
-        self.db_options.serialize_into_map(map);
-    }
-}
-
-impl ListingCatalogOptions {
-    pub fn builder() -> ListingCatalogOptionsBuilder {
-        ListingCatalogOptionsBuilder::new()
-    }
-
-    pub(crate) fn parse_from_map(map: &HashMap<String, String>) -> Result<Self> {
-        let db_options = ListingDatabaseOptions::parse_from_map(map)?;
-        Ok(Self { db_options })
-    }
-}
-
-#[derive(Clone, Debug, Default)]
-pub struct ListingCatalogOptionsBuilder {
-    options: ListingCatalogOptions,
-}
-
-impl ListingCatalogOptionsBuilder {
-    pub fn new() -> Self {
-        Self {
-            options: ListingCatalogOptions::default(),
-        }
-    }
-
-    pub fn db_options(mut self, db_options: ListingDatabaseOptions) -> Self {
-        self.options.db_options = db_options;
-        self
-    }
-
-    pub fn build(self) -> ListingCatalogOptions {
-        self.options
-    }
-}
-
-/// A catalog implementation that works by listing subfolders in a directory
-///
-/// The listing catalog will be created with a base folder specified by the URI.  Every subfolder
-/// in this base folder will be considered a database.  These will be opened as a
-/// [`crate::database::listing::ListingDatabase`]
-#[derive(Debug)]
-pub struct ListingCatalog {
-    object_store: Arc<ObjectStore>,
-
-    uri: String,
-
-    base_path: ObjectStorePath,
-
-    options: ListingCatalogOptions,
-}
-
-impl ListingCatalog {
-    /// Try to create a local directory to store the lancedb dataset
-    pub fn try_create_dir(path: &str) -> core::result::Result<(), std::io::Error> {
-        let path = Path::new(path);
-        if !path.try_exists()? {
-            create_dir_all(path)?;
-        }
-        Ok(())
-    }
-
-    pub fn uri(&self) -> &str {
-        &self.uri
-    }
-
-    async fn open_path(path: &str) -> Result<Self> {
-        let (object_store, base_path) = ObjectStore::from_uri(path).await?;
-        if object_store.is_local() {
-            Self::try_create_dir(path).context(CreateDirSnafu { path })?;
-        }
-
-        Ok(Self {
-            uri: path.to_string(),
-            base_path,
-            object_store,
-            options: ListingCatalogOptions::default(),
-        })
-    }
-
-    pub async fn connect(request: &ConnectRequest) -> Result<Self> {
-        let uri = &request.uri;
-        let parse_res = url::Url::parse(uri);
-
-        let options = ListingCatalogOptions::parse_from_map(&request.options)?;
-
-        match parse_res {
-            Ok(url) if url.scheme().len() == 1 && cfg!(windows) => Self::open_path(uri).await,
-            Ok(url) => {
-                let plain_uri = url.to_string();
-
-                let registry = Arc::new(ObjectStoreRegistry::default());
-                let storage_options = options.db_options.storage_options.clone();
-                let os_params = ObjectStoreParams {
-                    storage_options: Some(storage_options.clone()),
-                    ..Default::default()
-                };
-                let (object_store, base_path) =
-                    ObjectStore::from_uri_and_params(registry, &plain_uri, &os_params).await?;
-                if object_store.is_local() {
-                    Self::try_create_dir(&plain_uri).context(CreateDirSnafu { path: plain_uri })?;
-                }
-
-                Ok(Self {
-                    uri: String::from(url.clone()),
-                    base_path,
-                    object_store,
-                    options,
-                })
-            }
-            Err(_) => Self::open_path(uri).await,
-        }
-    }
-
-    fn database_path(&self, name: &str) -> ObjectStorePath {
-        self.base_path.child(name.replace('\\', "/"))
-    }
-}
-
-#[async_trait]
-impl Catalog for ListingCatalog {
-    async fn database_names(&self, request: DatabaseNamesRequest) -> Result<Vec<String>> {
-        let mut f = self
-            .object_store
-            .read_dir(self.base_path.clone())
-            .await?
-            .iter()
-            .map(Path::new)
-            .filter_map(|p| p.file_name().and_then(|s| s.to_str().map(String::from)))
-            .collect::<Vec<String>>();
-        f.sort();
-
-        if let Some(start_after) = request.start_after {
-            let index = f
-                .iter()
-                .position(|name| name.as_str() > start_after.as_str())
-                .unwrap_or(f.len());
-            f.drain(0..index);
-        }
-        if let Some(limit) = request.limit {
-            f.truncate(limit as usize);
-        }
-        Ok(f)
-    }
-
-    async fn create_database(&self, request: CreateDatabaseRequest) -> Result<Arc<dyn Database>> {
-        let db_path = self.database_path(&request.name);
-        let db_path_str = to_local_path(&db_path);
-        let exists = Path::new(&db_path_str).exists();
-
-        match request.mode {
-            CreateDatabaseMode::Create if exists => {
-                return Err(Error::DatabaseAlreadyExists { name: request.name })
-            }
-            CreateDatabaseMode::Create => {
-                create_dir_all(db_path.to_string()).unwrap();
-            }
-            CreateDatabaseMode::ExistOk => {
-                if !exists {
-                    create_dir_all(db_path.to_string()).unwrap();
-                }
-            }
-            CreateDatabaseMode::Overwrite => {
-                if exists {
-                    self.drop_database(&request.name).await?;
-                }
-                create_dir_all(db_path.to_string()).unwrap();
-            }
-        }
-
-        let db_uri = format!("/{}/{}", self.base_path, request.name);
-
-        let mut connect_request = ConnectRequest {
-            uri: db_uri,
-            #[cfg(feature = "remote")]
-            client_config: Default::default(),
-            read_consistency_interval: None,
-            options: Default::default(),
-            session: None,
-        };
-
-        // Add the db options to the connect request
-        self.options
-            .db_options
-            .serialize_into_map(&mut connect_request.options);
-
-        Ok(Arc::new(
-            ListingDatabase::connect_with_options(&connect_request).await?,
-        ))
-    }
-
-    async fn open_database(&self, request: OpenDatabaseRequest) -> Result<Arc<dyn Database>> {
-        let db_path = self.database_path(&request.name);
-
-        let db_path_str = to_local_path(&db_path);
-        let exists = Path::new(&db_path_str).exists();
-        if !exists {
-            return Err(Error::DatabaseNotFound { name: request.name });
-        }
-
-        let mut connect_request = ConnectRequest {
-            uri: db_path.to_string(),
-            #[cfg(feature = "remote")]
-            client_config: Default::default(),
-            read_consistency_interval: None,
-            options: Default::default(),
-            session: None,
-        };
-
-        // Add the db options to the connect request
-        self.options
-            .db_options
-            .serialize_into_map(&mut connect_request.options);
-
-        Ok(Arc::new(
-            ListingDatabase::connect_with_options(&connect_request).await?,
-        ))
-    }
-
-    async fn rename_database(&self, _old_name: &str, _new_name: &str) -> Result<()> {
-        Err(Error::NotSupported {
-            message: "rename_database is not supported in LanceDB OSS yet".to_string(),
-        })
-    }
-
-    async fn drop_database(&self, name: &str) -> Result<()> {
-        let db_path = self.database_path(name);
-        self.object_store
-            .remove_dir_all(db_path.clone())
-            .await
-            .map_err(|err| match err {
-                lance::Error::NotFound { .. } => Error::DatabaseNotFound {
-                    name: name.to_owned(),
-                },
-                _ => Error::from(err),
-            })?;
-
-        Ok(())
-    }
-
-    async fn drop_all_databases(&self) -> Result<()> {
-        self.object_store
-            .remove_dir_all(self.base_path.clone())
-            .await?;
-        Ok(())
-    }
-}
-
-#[cfg(all(test, not(windows)))]
-mod tests {
-    use super::*;
-
-    /// file:/// URIs with drive letters do not work correctly on Windows
-    #[cfg(windows)]
-    fn path_to_uri(path: PathBuf) -> String {
-        path.to_str().unwrap().to_string()
-    }
-
-    #[cfg(not(windows))]
-    fn path_to_uri(path: PathBuf) -> String {
-        Url::from_file_path(path).unwrap().to_string()
-    }
-
-    async fn setup_catalog() -> (TempDir, ListingCatalog) {
-        let tempdir = tempfile::tempdir().unwrap();
-        let catalog_path = tempdir.path().join("catalog");
-        std::fs::create_dir_all(&catalog_path).unwrap();
-
-        let uri = path_to_uri(catalog_path);
-
-        let request = ConnectRequest {
-            uri: uri.clone(),
-            #[cfg(feature = "remote")]
-            client_config: Default::default(),
-            options: Default::default(),
-            read_consistency_interval: None,
-            session: None,
-        };
-
-        let catalog = ListingCatalog::connect(&request).await.unwrap();
-
-        (tempdir, catalog)
-    }
-
-    use crate::database::{CreateTableData, CreateTableRequest, TableNamesRequest};
-    use crate::table::TableDefinition;
-    use arrow_schema::Field;
-    use std::path::PathBuf;
-    use std::sync::Arc;
-    use tempfile::{tempdir, TempDir};
-    use url::Url;
-
-    #[tokio::test]
-    async fn test_database_names() {
-        let (_tempdir, catalog) = setup_catalog().await;
-
-        let names = catalog
-            .database_names(DatabaseNamesRequest::default())
-            .await
-            .unwrap();
-        assert!(names.is_empty());
-    }
-
-    #[tokio::test]
-    async fn test_create_database() {
-        let (_tempdir, catalog) = setup_catalog().await;
-
-        catalog
-            .create_database(CreateDatabaseRequest {
-                name: "db1".into(),
-                mode: CreateDatabaseMode::Create,
-                options: HashMap::new(),
-            })
-            .await
-            .unwrap();
-
-        let names = catalog
-            .database_names(DatabaseNamesRequest::default())
-            .await
-            .unwrap();
-        assert_eq!(names, vec!["db1"]);
-    }
-
-    #[tokio::test]
-    async fn test_create_database_exist_ok() {
-        let (_tempdir, catalog) = setup_catalog().await;
-
-        let db1 = catalog
-            .create_database(CreateDatabaseRequest {
-                name: "db_exist_ok".into(),
-                mode: CreateDatabaseMode::ExistOk,
-                options: HashMap::new(),
-            })
-            .await
-            .unwrap();
-        let dummy_schema = Arc::new(arrow_schema::Schema::new(Vec::<Field>::default()));
-        db1.create_table(CreateTableRequest {
-            name: "test_table".parse().unwrap(),
-            data: CreateTableData::Empty(TableDefinition::new_from_schema(dummy_schema)),
-            mode: Default::default(),
-            write_options: Default::default(),
-        })
-        .await
-        .unwrap();
-
-        let db2 = catalog
-            .create_database(CreateDatabaseRequest {
-                name: "db_exist_ok".into(),
-                mode: CreateDatabaseMode::ExistOk,
-                options: HashMap::new(),
-            })
-            .await
-            .unwrap();
-
-        let tables = db2.table_names(TableNamesRequest::default()).await.unwrap();
-        assert_eq!(tables, vec!["test_table".to_string()]);
-    }
-
-    #[tokio::test]
-    async fn test_create_database_overwrite() {
-        let (_tempdir, catalog) = setup_catalog().await;
-
-        let db = catalog
-            .create_database(CreateDatabaseRequest {
-                name: "db_overwrite".into(),
-                mode: CreateDatabaseMode::Create,
-                options: HashMap::new(),
-            })
-            .await
-            .unwrap();
-        let dummy_schema = Arc::new(arrow_schema::Schema::new(Vec::<Field>::default()));
-        db.create_table(CreateTableRequest {
-            name: "old_table".parse().unwrap(),
-            data: CreateTableData::Empty(TableDefinition::new_from_schema(dummy_schema)),
-            mode: Default::default(),
-            write_options: Default::default(),
-        })
-        .await
-        .unwrap();
-        let tables = db.table_names(TableNamesRequest::default()).await.unwrap();
-        assert!(!tables.is_empty());
-
-        let new_db = catalog
-            .create_database(CreateDatabaseRequest {
-                name: "db_overwrite".into(),
-                mode: CreateDatabaseMode::Overwrite,
-                options: HashMap::new(),
-            })
-            .await
-            .unwrap();
-
-        let tables = new_db
-            .table_names(TableNamesRequest::default())
-            .await
-            .unwrap();
-        assert!(tables.is_empty());
-    }
-
-    #[tokio::test]
-    async fn test_create_database_overwrite_non_existing() {
-        let (_tempdir, catalog) = setup_catalog().await;
-
-        catalog
-            .create_database(CreateDatabaseRequest {
-                name: "new_db".into(),
-                mode: CreateDatabaseMode::Overwrite,
-                options: HashMap::new(),
-            })
-            .await
-            .unwrap();
-
-        let names = catalog
-            .database_names(DatabaseNamesRequest::default())
-            .await
-            .unwrap();
-        assert!(names.contains(&"new_db".to_string()));
-    }
-
-    #[tokio::test]
-    async fn test_open_database() {
-        let (_tempdir, catalog) = setup_catalog().await;
-
-        // Test open non-existent
-        let result = catalog
-            .open_database(OpenDatabaseRequest {
-                name: "missing".into(),
-                database_options: HashMap::new(),
-            })
-            .await;
-        assert!(matches!(
-            result.unwrap_err(),
-            Error::DatabaseNotFound { name } if name == "missing"
-        ));
-
-        // Create and open
-        catalog
-            .create_database(CreateDatabaseRequest {
-                name: "valid_db".into(),
-                mode: CreateDatabaseMode::Create,
-                options: HashMap::new(),
-            })
-            .await
-            .unwrap();
-
-        let db = catalog
-            .open_database(OpenDatabaseRequest {
-                name: "valid_db".into(),
-                database_options: HashMap::new(),
-            })
-            .await
-            .unwrap();
-        assert_eq!(
-            db.table_names(TableNamesRequest::default()).await.unwrap(),
-            Vec::<String>::new()
-        );
-    }
-
-    #[tokio::test]
-    async fn test_drop_database() {
-        let (_tempdir, catalog) = setup_catalog().await;
-
-        // Create test database
-        catalog
-            .create_database(CreateDatabaseRequest {
-                name: "to_drop".into(),
-                mode: CreateDatabaseMode::Create,
-                options: HashMap::new(),
-            })
-            .await
-            .unwrap();
-
-        let names = catalog
-            .database_names(DatabaseNamesRequest::default())
-            .await
-            .unwrap();
-        assert!(!names.is_empty());
-
-        // Drop database
-        catalog.drop_database("to_drop").await.unwrap();
-
-        let names = catalog
-            .database_names(DatabaseNamesRequest::default())
-            .await
-            .unwrap();
-        assert!(names.is_empty());
-    }
-
-    #[tokio::test]
-    async fn test_drop_all_databases() {
-        let (_tempdir, catalog) = setup_catalog().await;
-
-        catalog
-            .create_database(CreateDatabaseRequest {
-                name: "db1".into(),
-                mode: CreateDatabaseMode::Create,
-                options: HashMap::new(),
-            })
-            .await
-            .unwrap();
-        catalog
-            .create_database(CreateDatabaseRequest {
-                name: "db2".into(),
-                mode: CreateDatabaseMode::Create,
-                options: HashMap::new(),
-            })
-            .await
-            .unwrap();
-
-        catalog.drop_all_databases().await.unwrap();
-
-        let names = catalog
-            .database_names(DatabaseNamesRequest::default())
-            .await
-            .unwrap();
-        assert!(names.is_empty());
-    }
-
-    #[tokio::test]
-    async fn test_rename_database_unsupported() {
-        let (_tempdir, catalog) = setup_catalog().await;
-        let result = catalog.rename_database("old", "new").await;
-        assert!(matches!(
-            result.unwrap_err(),
-            Error::NotSupported { message } if message.contains("rename_database")
-        ));
-    }
-
-    #[tokio::test]
-    async fn test_connect_local_path() {
-        let tmp_dir = tempdir().unwrap();
-        let path = tmp_dir.path().to_str().unwrap();
-
-        let request = ConnectRequest {
-            uri: path.to_string(),
-            #[cfg(feature = "remote")]
-            client_config: Default::default(),
-            options: Default::default(),
-            read_consistency_interval: None,
-            session: None,
-        };
-
-        let catalog = ListingCatalog::connect(&request).await.unwrap();
-        assert!(catalog.object_store.is_local());
-        assert_eq!(catalog.uri, path);
-    }
-
-    #[tokio::test]
-    async fn test_connect_file_scheme() {
-        let tmp_dir = tempdir().unwrap();
-        let path = tmp_dir.path();
-        let uri = path_to_uri(path.to_path_buf());
-
-        let request = ConnectRequest {
-            uri: uri.clone(),
-            #[cfg(feature = "remote")]
-            client_config: Default::default(),
-            options: Default::default(),
-            read_consistency_interval: None,
-            session: None,
-        };
-
-        let catalog = ListingCatalog::connect(&request).await.unwrap();
-        assert!(catalog.object_store.is_local());
-        assert_eq!(catalog.uri, uri);
-    }
-
-    #[tokio::test]
-    async fn test_connect_invalid_uri_fallback() {
-        let invalid_uri = "invalid:///path";
-        let request = ConnectRequest {
-            uri: invalid_uri.to_string(),
-            #[cfg(feature = "remote")]
-            client_config: Default::default(),
-            options: Default::default(),
-            read_consistency_interval: None,
-            session: None,
-        };
-
-        let result = ListingCatalog::connect(&request).await;
-        assert!(result.is_err());
-    }
-}

rust/lancedb/src/connection.rs

@@ -9,16 +9,16 @@ use std::sync::Arc;
 use arrow_array::RecordBatchReader;
 use arrow_schema::{Field, SchemaRef};
 use lance::dataset::ReadParams;
+#[cfg(feature = "aws")]
 use object_store::aws::AwsCredential;
 
 use crate::arrow::{IntoArrow, IntoArrowStream, SendableRecordBatchStream};
-use crate::catalog::listing::ListingCatalog;
-use crate::catalog::CatalogOptions;
 use crate::database::listing::{
     ListingDatabase, OPT_NEW_TABLE_STORAGE_VERSION, OPT_NEW_TABLE_V2_MANIFEST_PATHS,
 };
 use crate::database::{
-    CreateTableData, CreateTableMode, CreateTableRequest, Database, DatabaseOptions,
+    CloneTableRequest, CreateNamespaceRequest, CreateTableData, CreateTableMode,
+    CreateTableRequest, Database, DatabaseOptions, DropNamespaceRequest, ListNamespacesRequest,
     OpenTableRequest, TableNamesRequest,
 };
 use crate::embeddings::{
@@ -66,6 +66,12 @@ impl TableNamesBuilder {
         self
     }
 
+    /// Set the namespace to list tables from
+    pub fn namespace(mut self, namespace: Vec<String>) -> Self {
+        self.request.namespace = namespace;
+        self
+    }
+
     /// Execute the table names operation
     pub async fn execute(self) -> Result<Vec<String>> {
         self.parent.clone().table_names(self.request).await
@@ -347,6 +353,12 @@ impl<const HAS_DATA: bool> CreateTableBuilder<HAS_DATA> {
         );
         self
     }
+
+    /// Set the namespace for the table
+    pub fn namespace(mut self, namespace: Vec<String>) -> Self {
+        self.request.namespace = namespace;
+        self
+    }
 }
 
 #[derive(Clone, Debug)]
@@ -366,6 +378,7 @@ impl OpenTableBuilder {
             parent,
             request: OpenTableRequest {
                 name,
+                namespace: vec![],
                 index_cache_size: None,
                 lance_read_params: None,
             },
@@ -441,6 +454,12 @@ impl OpenTableBuilder {
         self
     }
 
+    /// Set the namespace for the table
+    pub fn namespace(mut self, namespace: Vec<String>) -> Self {
+        self.request.namespace = namespace;
+        self
+    }
+
     /// Open the table
     pub async fn execute(self) -> Result<Table> {
         Ok(Table::new_with_embedding_registry(
@@ -450,6 +469,62 @@ impl OpenTableBuilder {
     }
 }
 
+/// Builder for cloning a table.
+///
+/// A shallow clone creates a new table that shares the underlying data files
+/// with the source table but has its own independent manifest. Both the source
+/// and cloned tables can evolve independently while initially sharing the same
+/// data, deletion, and index files.
+///
+/// Use this builder to configure the clone operation before executing it.
+pub struct CloneTableBuilder {
+    parent: Arc<dyn Database>,
+    request: CloneTableRequest,
+}
+
+impl CloneTableBuilder {
+    fn new(parent: Arc<dyn Database>, target_table_name: String, source_uri: String) -> Self {
+        Self {
+            parent,
+            request: CloneTableRequest::new(target_table_name, source_uri),
+        }
+    }
+
+    /// Set the source version to clone from
+    pub fn source_version(mut self, version: u64) -> Self {
+        self.request.source_version = Some(version);
+        self
+    }
+
+    /// Set the source tag to clone from
+    pub fn source_tag(mut self, tag: impl Into<String>) -> Self {
+        self.request.source_tag = Some(tag.into());
+        self
+    }
+
+    /// Set the target namespace for the cloned table
+    pub fn target_namespace(mut self, namespace: Vec<String>) -> Self {
+        self.request.target_namespace = namespace;
+        self
+    }
+
+    /// Set whether to perform a shallow clone (default: true)
+    ///
+    /// When true, the cloned table shares data files with the source table.
+    /// When false, performs a deep clone (not yet implemented).
+    pub fn is_shallow(mut self, is_shallow: bool) -> Self {
+        self.request.is_shallow = is_shallow;
+        self
+    }
+
+    /// Execute the clone operation
+    pub async fn execute(self) -> Result<Table> {
+        Ok(Table::new(
+            self.parent.clone().clone_table(self.request).await?,
+        ))
+    }
+}
+
 /// A connection to LanceDB
 #[derive(Clone)]
 pub struct Connection {
@@ -556,6 +631,30 @@ impl Connection {
         )
     }
 
+    /// Clone a table in the database
+    ///
+    /// Creates a new table by cloning from an existing source table.
+    /// By default, this performs a shallow clone where the new table shares
+    /// the underlying data files with the source table.
+    ///
+    /// # Parameters
+    /// - `target_table_name`: The name of the new table to create
+    /// - `source_uri`: The URI of the source table to clone from
+    ///
+    /// # Returns
+    /// A [`CloneTableBuilder`] that can be used to configure the clone operation
+    pub fn clone_table(
+        &self,
+        target_table_name: impl Into<String>,
+        source_uri: impl Into<String>,
+    ) -> CloneTableBuilder {
+        CloneTableBuilder::new(
+            self.internal.clone(),
+            target_table_name.into(),
+            source_uri.into(),
+        )
+    }
+
     /// Rename a table in the database.
     ///
     /// This is only supported in LanceDB Cloud.
@@ -563,9 +662,16 @@ impl Connection {
         &self,
         old_name: impl AsRef<str>,
         new_name: impl AsRef<str>,
+        cur_namespace: &[String],
+        new_namespace: &[String],
     ) -> Result<()> {
         self.internal
-            .rename_table(old_name.as_ref(), new_name.as_ref())
+            .rename_table(
+                old_name.as_ref(),
+                new_name.as_ref(),
+                cur_namespace,
+                new_namespace,
+            )
             .await
     }
 
@@ -573,8 +679,9 @@ impl Connection {
     ///
     /// # Arguments
     /// * `name` - The name of the table to drop
-    pub async fn drop_table(&self, name: impl AsRef<str>) -> Result<()> {
-        self.internal.drop_table(name.as_ref()).await
+    /// * `namespace` - The namespace to drop the table from
+    pub async fn drop_table(&self, name: impl AsRef<str>, namespace: &[String]) -> Result<()> {
+        self.internal.drop_table(name.as_ref(), namespace).await
     }
 
     /// Drop the database
@@ -582,12 +689,30 @@ impl Connection {
     /// This is the same as dropping all of the tables
     #[deprecated(since = "0.15.1", note = "Use `drop_all_tables` instead")]
     pub async fn drop_db(&self) -> Result<()> {
-        self.internal.drop_all_tables().await
+        self.internal.drop_all_tables(&[]).await
     }
 
     /// Drops all tables in the database
-    pub async fn drop_all_tables(&self) -> Result<()> {
-        self.internal.drop_all_tables().await
+    ///
+    /// # Arguments
+    /// * `namespace` - The namespace to drop all tables from. Empty slice represents root namespace.
+    pub async fn drop_all_tables(&self, namespace: &[String]) -> Result<()> {
+        self.internal.drop_all_tables(namespace).await
+    }
+
+    /// List immediate child namespace names in the given namespace
+    pub async fn list_namespaces(&self, request: ListNamespacesRequest) -> Result<Vec<String>> {
+        self.internal.list_namespaces(request).await
+    }
+
+    /// Create a new namespace
+    pub async fn create_namespace(&self, request: CreateNamespaceRequest) -> Result<()> {
+        self.internal.create_namespace(request).await
+    }
+
+    /// Drop a namespace
+    pub async fn drop_namespace(&self, request: DropNamespaceRequest) -> Result<()> {
+        self.internal.drop_namespace(request).await
     }
 
     /// Get the in-memory embedding registry.
@@ -613,7 +738,7 @@ pub struct ConnectRequest {
     #[cfg(feature = "remote")]
     pub client_config: ClientConfig,
 
-    /// Database/Catalog specific options
+    /// Database specific options
     pub options: HashMap<String, String>,
 
     /// The interval at which to check for updates from other processes.
@@ -749,6 +874,7 @@ impl ConnectBuilder {
     }
 
     /// [`AwsCredential`] to use when connecting to S3.
+    #[cfg(feature = "aws")]
     #[deprecated(note = "Pass through storage_options instead")]
     pub fn aws_creds(mut self, aws_creds: AwsCredential) -> Self {
         self.request
@@ -889,50 +1015,6 @@ pub fn connect(uri: &str) -> ConnectBuilder {
     ConnectBuilder::new(uri)
 }
 
-/// A builder for configuring a connection to a LanceDB catalog
-#[derive(Debug)]
-pub struct CatalogConnectBuilder {
-    request: ConnectRequest,
-}
-
-impl CatalogConnectBuilder {
-    /// Create a new [`CatalogConnectBuilder`] with the given catalog URI.
-    pub fn new(uri: &str) -> Self {
-        Self {
-            request: ConnectRequest {
-                uri: uri.to_string(),
-                #[cfg(feature = "remote")]
-                client_config: Default::default(),
-                read_consistency_interval: None,
-                options: HashMap::new(),
-                session: None,
-            },
-        }
-    }
-
-    pub fn catalog_options(mut self, catalog_options: &dyn CatalogOptions) -> Self {
-        catalog_options.serialize_into_map(&mut self.request.options);
-        self
-    }
-
-    /// Establishes a connection to the catalog
-    pub async fn execute(self) -> Result<Arc<ListingCatalog>> {
-        let catalog = ListingCatalog::connect(&self.request).await?;
-        Ok(Arc::new(catalog))
-    }
-}
-
-/// Connect to a LanceDB catalog.
-///
-/// A catalog is a container for databases, which in turn are containers for tables.
-///
-/// # Arguments
-///
-/// * `uri` - URI where the catalog is located, can be a local directory or supported remote cloud storage.
-pub fn connect_catalog(uri: &str) -> CatalogConnectBuilder {
-    CatalogConnectBuilder::new(uri)
-}
-
 #[cfg(all(test, feature = "remote"))]
 mod test_utils {
     use super::*;
@@ -950,6 +1032,23 @@ mod test_utils {
                 embedding_registry: Arc::new(MemoryRegistry::new()),
             }
         }
+
+        pub fn new_with_handler_and_config<T>(
+            handler: impl Fn(reqwest::Request) -> http::Response<T> + Clone + Send + Sync + 'static,
+            config: crate::remote::ClientConfig,
+        ) -> Self
+        where
+            T: Into<reqwest::Body>,
+        {
+            let internal = Arc::new(crate::remote::db::RemoteDatabase::new_mock_with_config(
+                handler, config,
+            ));
+            Self {
+                internal,
+                uri: "db://test".to_string(),
+                embedding_registry: Arc::new(MemoryRegistry::new()),
+            }
+        }
     }
 }
 
@@ -957,7 +1056,6 @@ mod test_utils {
 mod tests {
     use std::fs::create_dir_all;
 
-    use crate::catalog::{Catalog, DatabaseNamesRequest, OpenDatabaseRequest};
     use crate::database::listing::{ListingDatabaseOptions, NewTableConfig};
     use crate::query::QueryBase;
     use crate::query::{ExecutableQuery, QueryExecutionOptions};
@@ -1218,12 +1316,12 @@ mod tests {
 
         // drop non-exist table
         assert!(matches!(
-            db.drop_table("invalid_table").await,
+            db.drop_table("invalid_table", &[]).await,
             Err(crate::Error::TableNotFound { .. }),
         ));
 
         create_dir_all(tmp_dir.path().join("table1.lance")).unwrap();
-        db.drop_table("table1").await.unwrap();
+        db.drop_table("table1", &[]).await.unwrap();
 
         let tables = db.table_names().execute().await.unwrap();
         assert_eq!(tables.len(), 0);
@@ -1265,89 +1363,48 @@ mod tests {
     }
 
     #[tokio::test]
-    async fn test_connect_catalog() {
+    async fn test_clone_table() {
         let tmp_dir = tempdir().unwrap();
         let uri = tmp_dir.path().to_str().unwrap();
-        let catalog = connect_catalog(uri).execute().await.unwrap();
+        let db = connect(uri).execute().await.unwrap();
 
-        // Verify that we can get the uri from the catalog
-        let catalog_uri = catalog.uri();
-        assert_eq!(catalog_uri, uri);
+        // Create a source table with some data
+        let mut batch_gen = BatchGenerator::new()
+            .col(Box::new(IncrementingInt32::new().named("id")))
+            .col(Box::new(IncrementingInt32::new().named("value")));
+        let reader = batch_gen.batches(5, 100);
 
-        // Check that the catalog is initially empty
-        let dbs = catalog
-            .database_names(DatabaseNamesRequest::default())
-            .await
-            .unwrap();
-        assert_eq!(dbs.len(), 0);
-    }
-
-    #[tokio::test]
-    #[cfg(not(windows))]
-    async fn test_catalog_create_database() {
-        let tmp_dir = tempdir().unwrap();
-        let uri = tmp_dir.path().to_str().unwrap();
-        let catalog = connect_catalog(uri).execute().await.unwrap();
-
-        let db_name = "test_db";
-        catalog
-            .create_database(crate::catalog::CreateDatabaseRequest {
-                name: db_name.to_string(),
-                mode: Default::default(),
-                options: Default::default(),
-            })
+        let source_table = db
+            .create_table("source_table", reader)
+            .execute()
             .await
             .unwrap();
 
-        let dbs = catalog
-            .database_names(DatabaseNamesRequest::default())
-            .await
-            .unwrap();
-        assert_eq!(dbs.len(), 1);
-        assert_eq!(dbs[0], db_name);
+        // Get the source table URI
+        let source_table_path = tmp_dir.path().join("source_table.lance");
+        let source_uri = source_table_path.to_str().unwrap();
 
-        let db = catalog
-            .open_database(OpenDatabaseRequest {
-                name: db_name.to_string(),
-                database_options: HashMap::new(),
-            })
+        // Clone the table
+        let cloned_table = db
+            .clone_table("cloned_table", source_uri)
+            .execute()
             .await
             .unwrap();
 
-        let tables = db.table_names(Default::default()).await.unwrap();
-        assert_eq!(tables.len(), 0);
-    }
+        // Verify the cloned table exists
+        let table_names = db.table_names().execute().await.unwrap();
+        assert!(table_names.contains(&"source_table".to_string()));
+        assert!(table_names.contains(&"cloned_table".to_string()));
 
-    #[tokio::test]
-    #[cfg(not(windows))]
-    async fn test_catalog_drop_database() {
-        let tmp_dir = tempdir().unwrap();
-        let uri = tmp_dir.path().to_str().unwrap();
-        let catalog = connect_catalog(uri).execute().await.unwrap();
+        // Verify the cloned table has the same schema
+        assert_eq!(
+            source_table.schema().await.unwrap(),
+            cloned_table.schema().await.unwrap()
+        );
 
-        // Create and then drop a database
-        let db_name = "test_db_to_drop";
-        catalog
-            .create_database(crate::catalog::CreateDatabaseRequest {
-                name: db_name.to_string(),
-                mode: Default::default(),
-                options: Default::default(),
-            })
-            .await
-            .unwrap();
-
-        let dbs = catalog
-            .database_names(DatabaseNamesRequest::default())
-            .await
-            .unwrap();
-        assert_eq!(dbs.len(), 1);
-
-        catalog.drop_database(db_name).await.unwrap();
-
-        let dbs_after = catalog
-            .database_names(DatabaseNamesRequest::default())
-            .await
-            .unwrap();
-        assert_eq!(dbs_after.len(), 0);
+        // Verify the cloned table has the same data
+        let source_count = source_table.count_rows(None).await.unwrap();
+        let cloned_count = cloned_table.count_rows(None).await.unwrap();
+        assert_eq!(source_count, cloned_count);
     }
 }

rust/lancedb/src/database.rs

@@ -34,9 +34,36 @@ pub trait DatabaseOptions {
     fn serialize_into_map(&self, map: &mut HashMap<String, String>);
 }
 
+/// A request to list namespaces in the database
+#[derive(Clone, Debug, Default)]
+pub struct ListNamespacesRequest {
+    /// The parent namespace to list namespaces in. Empty list represents root namespace.
+    pub namespace: Vec<String>,
+    /// If present, only return names that come lexicographically after the supplied value.
+    pub page_token: Option<String>,
+    /// The maximum number of namespace names to return
+    pub limit: Option<u32>,
+}
+
+/// A request to create a namespace
+#[derive(Clone, Debug)]
+pub struct CreateNamespaceRequest {
+    /// The namespace identifier to create
+    pub namespace: Vec<String>,
+}
+
+/// A request to drop a namespace
+#[derive(Clone, Debug)]
+pub struct DropNamespaceRequest {
+    /// The namespace identifier to drop
+    pub namespace: Vec<String>,
+}
+
 /// A request to list names of tables in the database
 #[derive(Clone, Debug, Default)]
 pub struct TableNamesRequest {
+    /// The namespace to list tables in. Empty list represents root namespace.
+    pub namespace: Vec<String>,
     /// If present, only return names that come lexicographically after the supplied
     /// value.
     ///
@@ -51,6 +78,8 @@ pub struct TableNamesRequest {
 #[derive(Clone, Debug)]
 pub struct OpenTableRequest {
     pub name: String,
+    /// The namespace to open the table from. Empty list represents root namespace.
+    pub namespace: Vec<String>,
     pub index_cache_size: Option<u32>,
     pub lance_read_params: Option<ReadParams>,
 }
@@ -125,6 +154,8 @@ impl StreamingWriteSource for CreateTableData {
 pub struct CreateTableRequest {
     /// The name of the new table
     pub name: String,
+    /// The namespace to create the table in. Empty list represents root namespace.
+    pub namespace: Vec<String>,
     /// Initial data to write to the table, can be None to create an empty table
     pub data: CreateTableData,
     /// The mode to use when creating the table
@@ -137,6 +168,7 @@ impl CreateTableRequest {
     pub fn new(name: String, data: CreateTableData) -> Self {
         Self {
             name,
+            namespace: vec![],
             data,
             mode: CreateTableMode::default(),
             write_options: WriteOptions::default(),
@@ -144,6 +176,42 @@ impl CreateTableRequest {
     }
 }
 
+/// Request to 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.
+#[derive(Clone, Debug)]
+pub struct CloneTableRequest {
+    /// The name of the target table to create
+    pub target_table_name: String,
+    /// The namespace for the target table. Empty list represents root namespace.
+    pub target_namespace: Vec<String>,
+    /// The URI of the source table to clone from.
+    pub source_uri: String,
+    /// Optional version of the source table to clone.
+    pub source_version: Option<u64>,
+    /// Optional tag of the source table to clone.
+    pub source_tag: Option<String>,
+    /// Whether to perform a shallow clone (true) or deep clone (false). Defaults to true.
+    /// Currently only shallow clone is supported.
+    pub is_shallow: bool,
+}
+
+impl CloneTableRequest {
+    pub fn new(target_table_name: String, source_uri: String) -> Self {
+        Self {
+            target_table_name,
+            target_namespace: vec![],
+            source_uri,
+            source_version: None,
+            source_tag: None,
+            is_shallow: true,
+        }
+    }
+}
+
 /// The `Database` trait defines the interface for database implementations.
 ///
 /// A database is responsible for managing tables and their metadata.
@@ -151,17 +219,36 @@ impl CreateTableRequest {
 pub trait Database:
     Send + Sync + std::any::Any + std::fmt::Debug + std::fmt::Display + 'static
 {
+    /// List immediate child namespace names in the given namespace
+    async fn list_namespaces(&self, request: ListNamespacesRequest) -> Result<Vec<String>>;
+    /// Create a new namespace
+    async fn create_namespace(&self, request: CreateNamespaceRequest) -> Result<()>;
+    /// Drop a namespace
+    async fn drop_namespace(&self, request: DropNamespaceRequest) -> Result<()>;
     /// List the names of tables in the database
     async fn table_names(&self, request: TableNamesRequest) -> Result<Vec<String>>;
     /// Create a table in the database
     async fn create_table(&self, request: CreateTableRequest) -> Result<Arc<dyn BaseTable>>;
+    /// Clone a table in the database.
+    ///
+    /// Creates a shallow clone of the source table, sharing underlying data files
+    /// but with an independent manifest. Both tables can evolve separately after cloning.
+    ///
+    /// See [`CloneTableRequest`] for detailed documentation and examples.
+    async fn clone_table(&self, request: CloneTableRequest) -> Result<Arc<dyn BaseTable>>;
     /// Open a table in the database
     async fn open_table(&self, request: OpenTableRequest) -> Result<Arc<dyn BaseTable>>;
     /// Rename a table in the database
-    async fn rename_table(&self, old_name: &str, new_name: &str) -> Result<()>;
+    async fn rename_table(
+        &self,
+        cur_name: &str,
+        new_name: &str,
+        cur_namespace: &[String],
+        new_namespace: &[String],
+    ) -> Result<()>;
     /// Drop a table in the database
-    async fn drop_table(&self, name: &str) -> Result<()>;
+    async fn drop_table(&self, name: &str, namespace: &[String]) -> Result<()>;
     /// Drop all tables in the database
-    async fn drop_all_tables(&self) -> Result<()>;
+    async fn drop_all_tables(&self, namespace: &[String]) -> Result<()>;
     fn as_any(&self) -> &dyn std::any::Any;
 }

rust/lancedb/src/database/listing.rs

@@ -7,7 +7,8 @@ use std::fs::create_dir_all;
 use std::path::Path;
 use std::{collections::HashMap, sync::Arc};
 
-use lance::dataset::{ReadParams, WriteMode};
+use lance::dataset::refs::Ref;
+use lance::dataset::{builder::DatasetBuilder, ReadParams, WriteMode};
 use lance::io::{ObjectStore, ObjectStoreParams, WrappingObjectStore};
 use lance_datafusion::utils::StreamingWriteSource;
 use lance_encoding::version::LanceFileVersion;
@@ -22,7 +23,8 @@ use crate::table::NativeTable;
 use crate::utils::validate_table_name;
 
 use super::{
-    BaseTable, CreateTableMode, CreateTableRequest, Database, DatabaseOptions, OpenTableRequest,
+    BaseTable, CloneTableRequest, CreateNamespaceRequest, CreateTableMode, CreateTableRequest,
+    Database, DatabaseOptions, DropNamespaceRequest, ListNamespacesRequest, OpenTableRequest,
     TableNamesRequest,
 };
 
@@ -551,6 +553,7 @@ impl ListingDatabase {
     async fn handle_table_exists(
         &self,
         table_name: &str,
+        namespace: Vec<String>,
         mode: CreateTableMode,
         data_schema: &arrow_schema::Schema,
     ) -> Result<Arc<dyn BaseTable>> {
@@ -561,6 +564,7 @@ impl ListingDatabase {
             CreateTableMode::ExistOk(callback) => {
                 let req = OpenTableRequest {
                     name: table_name.to_string(),
+                    namespace: namespace.clone(),
                     index_cache_size: None,
                     lance_read_params: None,
                 };
@@ -584,7 +588,34 @@ impl ListingDatabase {
 
 #[async_trait::async_trait]
 impl Database for ListingDatabase {
+    async fn list_namespaces(&self, request: ListNamespacesRequest) -> Result<Vec<String>> {
+        if !request.namespace.is_empty() {
+            return Err(Error::NotSupported {
+                message: "Namespace operations are not supported for listing database".into(),
+            });
+        }
+
+        Ok(Vec::new())
+    }
+
+    async fn create_namespace(&self, _request: CreateNamespaceRequest) -> Result<()> {
+        Err(Error::NotSupported {
+            message: "Namespace operations are not supported for listing database".into(),
+        })
+    }
+
+    async fn drop_namespace(&self, _request: DropNamespaceRequest) -> Result<()> {
+        Err(Error::NotSupported {
+            message: "Namespace operations are not supported for listing database".into(),
+        })
+    }
+
     async fn table_names(&self, request: TableNamesRequest) -> Result<Vec<String>> {
+        if !request.namespace.is_empty() {
+            return Err(Error::NotSupported {
+                message: "Namespace parameter is not supported for listing database. Only root namespace is supported.".into(),
+            });
+        }
         let mut f = self
             .object_store
             .read_dir(self.base_path.clone())
@@ -615,6 +646,11 @@ impl Database for ListingDatabase {
     }
 
     async fn create_table(&self, request: CreateTableRequest) -> Result<Arc<dyn BaseTable>> {
+        if !request.namespace.is_empty() {
+            return Err(Error::NotSupported {
+                message: "Namespace parameter is not supported for listing database. Only root namespace is supported.".into(),
+            });
+        }
         let table_uri = self.table_uri(&request.name)?;
 
         let (storage_version_override, v2_manifest_override) =
@@ -637,14 +673,83 @@ impl Database for ListingDatabase {
         {
             Ok(table) => Ok(Arc::new(table)),
             Err(Error::TableAlreadyExists { .. }) => {
-                self.handle_table_exists(&request.name, request.mode, &data_schema)
-                    .await
+                self.handle_table_exists(
+                    &request.name,
+                    request.namespace.clone(),
+                    request.mode,
+                    &data_schema,
+                )
+                .await
             }
             Err(err) => Err(err),
         }
     }
 
+    async fn clone_table(&self, request: CloneTableRequest) -> Result<Arc<dyn BaseTable>> {
+        if !request.target_namespace.is_empty() {
+            return Err(Error::NotSupported {
+                message: "Namespace parameter is not supported for listing database. Only root namespace is supported.".into(),
+            });
+        }
+
+        // TODO: support deep clone
+        if !request.is_shallow {
+            return Err(Error::NotSupported {
+                message: "Deep clone is not yet implemented".to_string(),
+            });
+        }
+
+        validate_table_name(&request.target_table_name)?;
+
+        let storage_params = ObjectStoreParams {
+            storage_options: Some(self.storage_options.clone()),
+            ..Default::default()
+        };
+        let read_params = ReadParams {
+            store_options: Some(storage_params.clone()),
+            session: Some(self.session.clone()),
+            ..Default::default()
+        };
+
+        let mut source_dataset = DatasetBuilder::from_uri(&request.source_uri)
+            .with_read_params(read_params.clone())
+            .load()
+            .await
+            .map_err(|e| Error::Lance { source: e })?;
+
+        let version_ref = match (request.source_version, request.source_tag) {
+            (Some(v), None) => Ok(Ref::Version(v)),
+            (None, Some(tag)) => Ok(Ref::Tag(tag)),
+            (None, None) => Ok(Ref::Version(source_dataset.version().version)),
+            _ => Err(Error::InvalidInput {
+                message: "Cannot specify both source_version and source_tag".to_string(),
+            }),
+        }?;
+
+        let target_uri = self.table_uri(&request.target_table_name)?;
+        source_dataset
+            .shallow_clone(&target_uri, version_ref, storage_params)
+            .await
+            .map_err(|e| Error::Lance { source: e })?;
+
+        let cloned_table = NativeTable::open_with_params(
+            &target_uri,
+            &request.target_table_name,
+            self.store_wrapper.clone(),
+            None,
+            self.read_consistency_interval,
+        )
+        .await?;
+
+        Ok(Arc::new(cloned_table))
+    }
+
     async fn open_table(&self, mut request: OpenTableRequest) -> Result<Arc<dyn BaseTable>> {
+        if !request.namespace.is_empty() {
+            return Err(Error::NotSupported {
+                message: "Namespace parameter is not supported for listing database. Only root namespace is supported.".into(),
+            });
+        }
         let table_uri = self.table_uri(&request.name)?;
 
         // Only modify the storage options if we actually have something to
@@ -694,17 +799,44 @@ impl Database for ListingDatabase {
         Ok(native_table)
     }
 
-    async fn rename_table(&self, _old_name: &str, _new_name: &str) -> Result<()> {
+    async fn rename_table(
+        &self,
+        _cur_name: &str,
+        _new_name: &str,
+        cur_namespace: &[String],
+        new_namespace: &[String],
+    ) -> Result<()> {
+        if !cur_namespace.is_empty() {
+            return Err(Error::NotSupported {
+                message: "Namespace parameter is not supported for listing database.".into(),
+            });
+        }
+        if !new_namespace.is_empty() {
+            return Err(Error::NotSupported {
+                message: "Namespace parameter is not supported for listing database.".into(),
+            });
+        }
         Err(Error::NotSupported {
-            message: "rename_table is not supported in LanceDB OSS".to_string(),
+            message: "rename_table is not supported in LanceDB OSS".into(),
         })
     }
 
-    async fn drop_table(&self, name: &str) -> Result<()> {
+    async fn drop_table(&self, name: &str, namespace: &[String]) -> Result<()> {
+        if !namespace.is_empty() {
+            return Err(Error::NotSupported {
+                message: "Namespace parameter is not supported for listing database.".into(),
+            });
+        }
         self.drop_tables(vec![name.to_string()]).await
     }
 
-    async fn drop_all_tables(&self) -> Result<()> {
+    async fn drop_all_tables(&self, namespace: &[String]) -> Result<()> {
+        // Check if namespace parameter is provided
+        if !namespace.is_empty() {
+            return Err(Error::NotSupported {
+                message: "Namespace parameter is not supported for listing database.".into(),
+            });
+        }
         let tables = self.table_names(TableNamesRequest::default()).await?;
         self.drop_tables(tables).await
     }
@@ -713,3 +845,694 @@ impl Database for ListingDatabase {
         self
     }
 }
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use crate::connection::ConnectRequest;
+    use crate::database::{CreateTableData, CreateTableMode, CreateTableRequest};
+    use crate::table::{Table, TableDefinition};
+    use arrow_array::{Int32Array, RecordBatch, StringArray};
+    use arrow_schema::{DataType, Field, Schema};
+    use tempfile::tempdir;
+
+    async fn setup_database() -> (tempfile::TempDir, ListingDatabase) {
+        let tempdir = tempdir().unwrap();
+        let uri = tempdir.path().to_str().unwrap();
+
+        let request = ConnectRequest {
+            uri: uri.to_string(),
+            #[cfg(feature = "remote")]
+            client_config: Default::default(),
+            options: Default::default(),
+            read_consistency_interval: None,
+            session: None,
+        };
+
+        let db = ListingDatabase::connect_with_options(&request)
+            .await
+            .unwrap();
+
+        (tempdir, db)
+    }
+
+    #[tokio::test]
+    async fn test_clone_table_basic() {
+        let (_tempdir, db) = setup_database().await;
+
+        // Create a source table with schema
+        let schema = Arc::new(Schema::new(vec![
+            Field::new("id", DataType::Int32, false),
+            Field::new("name", DataType::Utf8, false),
+        ]));
+
+        let source_table = db
+            .create_table(CreateTableRequest {
+                name: "source_table".to_string(),
+                namespace: vec![],
+                data: CreateTableData::Empty(TableDefinition::new_from_schema(schema.clone())),
+                mode: CreateTableMode::Create,
+                write_options: Default::default(),
+            })
+            .await
+            .unwrap();
+
+        // Get the source table URI
+        let source_uri = db.table_uri("source_table").unwrap();
+
+        // Clone the table
+        let cloned_table = db
+            .clone_table(CloneTableRequest {
+                target_table_name: "cloned_table".to_string(),
+                target_namespace: vec![],
+                source_uri: source_uri.clone(),
+                source_version: None,
+                source_tag: None,
+                is_shallow: true,
+            })
+            .await
+            .unwrap();
+
+        // Verify both tables exist
+        let table_names = db.table_names(TableNamesRequest::default()).await.unwrap();
+        assert!(table_names.contains(&"source_table".to_string()));
+        assert!(table_names.contains(&"cloned_table".to_string()));
+
+        // Verify schemas match
+        assert_eq!(
+            source_table.schema().await.unwrap(),
+            cloned_table.schema().await.unwrap()
+        );
+    }
+
+    #[tokio::test]
+    async fn test_clone_table_with_data() {
+        let (_tempdir, db) = setup_database().await;
+
+        // Create a source table with actual data
+        let schema = Arc::new(Schema::new(vec![
+            Field::new("id", DataType::Int32, false),
+            Field::new("name", DataType::Utf8, false),
+        ]));
+
+        let batch = RecordBatch::try_new(
+            schema.clone(),
+            vec![
+                Arc::new(Int32Array::from(vec![1, 2, 3])),
+                Arc::new(StringArray::from(vec!["a", "b", "c"])),
+            ],
+        )
+        .unwrap();
+
+        let reader = Box::new(arrow_array::RecordBatchIterator::new(
+            vec![Ok(batch)],
+            schema.clone(),
+        ));
+
+        let source_table = db
+            .create_table(CreateTableRequest {
+                name: "source_with_data".to_string(),
+                namespace: vec![],
+                data: CreateTableData::Data(reader),
+                mode: CreateTableMode::Create,
+                write_options: Default::default(),
+            })
+            .await
+            .unwrap();
+
+        let source_uri = db.table_uri("source_with_data").unwrap();
+
+        // Clone the table
+        let cloned_table = db
+            .clone_table(CloneTableRequest {
+                target_table_name: "cloned_with_data".to_string(),
+                target_namespace: vec![],
+                source_uri,
+                source_version: None,
+                source_tag: None,
+                is_shallow: true,
+            })
+            .await
+            .unwrap();
+
+        // Verify data counts match
+        let source_count = source_table.count_rows(None).await.unwrap();
+        let cloned_count = cloned_table.count_rows(None).await.unwrap();
+        assert_eq!(source_count, cloned_count);
+        assert_eq!(source_count, 3);
+    }
+
+    #[tokio::test]
+    async fn test_clone_table_with_storage_options() {
+        let tempdir = tempdir().unwrap();
+        let uri = tempdir.path().to_str().unwrap();
+
+        // Create database with storage options
+        let mut options = HashMap::new();
+        options.insert("test_option".to_string(), "test_value".to_string());
+
+        let request = ConnectRequest {
+            uri: uri.to_string(),
+            #[cfg(feature = "remote")]
+            client_config: Default::default(),
+            options: options.clone(),
+            read_consistency_interval: None,
+            session: None,
+        };
+
+        let db = ListingDatabase::connect_with_options(&request)
+            .await
+            .unwrap();
+
+        // Create source table
+        let schema = Arc::new(Schema::new(vec![Field::new("id", DataType::Int32, false)]));
+
+        db.create_table(CreateTableRequest {
+            name: "source".to_string(),
+            namespace: vec![],
+            data: CreateTableData::Empty(TableDefinition::new_from_schema(schema)),
+            mode: CreateTableMode::Create,
+            write_options: Default::default(),
+        })
+        .await
+        .unwrap();
+
+        let source_uri = db.table_uri("source").unwrap();
+
+        // Clone should work with storage options
+        let cloned = db
+            .clone_table(CloneTableRequest {
+                target_table_name: "cloned".to_string(),
+                target_namespace: vec![],
+                source_uri,
+                source_version: None,
+                source_tag: None,
+                is_shallow: true,
+            })
+            .await;
+
+        assert!(cloned.is_ok());
+    }
+
+    #[tokio::test]
+    async fn test_clone_table_deep_not_supported() {
+        let (_tempdir, db) = setup_database().await;
+
+        // Create a source table
+        let schema = Arc::new(Schema::new(vec![Field::new("id", DataType::Int32, false)]));
+
+        db.create_table(CreateTableRequest {
+            name: "source".to_string(),
+            namespace: vec![],
+            data: CreateTableData::Empty(TableDefinition::new_from_schema(schema)),
+            mode: CreateTableMode::Create,
+            write_options: Default::default(),
+        })
+        .await
+        .unwrap();
+
+        let source_uri = db.table_uri("source").unwrap();
+
+        // Try deep clone (should fail)
+        let result = db
+            .clone_table(CloneTableRequest {
+                target_table_name: "cloned".to_string(),
+                target_namespace: vec![],
+                source_uri,
+                source_version: None,
+                source_tag: None,
+                is_shallow: false, // Request deep clone
+            })
+            .await;
+
+        assert!(result.is_err());
+        assert!(matches!(
+            result.unwrap_err(),
+            Error::NotSupported { message } if message.contains("Deep clone")
+        ));
+    }
+
+    #[tokio::test]
+    async fn test_clone_table_with_namespace_not_supported() {
+        let (_tempdir, db) = setup_database().await;
+
+        // Create a source table
+        let schema = Arc::new(Schema::new(vec![Field::new("id", DataType::Int32, false)]));
+
+        db.create_table(CreateTableRequest {
+            name: "source".to_string(),
+            namespace: vec![],
+            data: CreateTableData::Empty(TableDefinition::new_from_schema(schema)),
+            mode: CreateTableMode::Create,
+            write_options: Default::default(),
+        })
+        .await
+        .unwrap();
+
+        let source_uri = db.table_uri("source").unwrap();
+
+        // Try clone with namespace (should fail for listing database)
+        let result = db
+            .clone_table(CloneTableRequest {
+                target_table_name: "cloned".to_string(),
+                target_namespace: vec!["namespace".to_string()], // Non-empty namespace
+                source_uri,
+                source_version: None,
+                source_tag: None,
+                is_shallow: true,
+            })
+            .await;
+
+        assert!(result.is_err());
+        assert!(matches!(
+            result.unwrap_err(),
+            Error::NotSupported { message } if message.contains("Namespace parameter is not supported")
+        ));
+    }
+
+    #[tokio::test]
+    async fn test_clone_table_invalid_target_name() {
+        let (_tempdir, db) = setup_database().await;
+
+        // Create a source table
+        let schema = Arc::new(Schema::new(vec![Field::new("id", DataType::Int32, false)]));
+
+        db.create_table(CreateTableRequest {
+            name: "source".to_string(),
+            namespace: vec![],
+            data: CreateTableData::Empty(TableDefinition::new_from_schema(schema)),
+            mode: CreateTableMode::Create,
+            write_options: Default::default(),
+        })
+        .await
+        .unwrap();
+
+        let source_uri = db.table_uri("source").unwrap();
+
+        // Try clone with invalid target name
+        let result = db
+            .clone_table(CloneTableRequest {
+                target_table_name: "invalid/name".to_string(), // Invalid name with slash
+                target_namespace: vec![],
+                source_uri,
+                source_version: None,
+                source_tag: None,
+                is_shallow: true,
+            })
+            .await;
+
+        assert!(result.is_err());
+    }
+
+    #[tokio::test]
+    async fn test_clone_table_source_not_found() {
+        let (_tempdir, db) = setup_database().await;
+
+        // Try to clone from non-existent source
+        let result = db
+            .clone_table(CloneTableRequest {
+                target_table_name: "cloned".to_string(),
+                target_namespace: vec![],
+                source_uri: "/nonexistent/table.lance".to_string(),
+                source_version: None,
+                source_tag: None,
+                is_shallow: true,
+            })
+            .await;
+
+        assert!(result.is_err());
+    }
+
+    #[tokio::test]
+    async fn test_clone_table_with_version_and_tag_error() {
+        let (_tempdir, db) = setup_database().await;
+
+        // Create a source table
+        let schema = Arc::new(Schema::new(vec![Field::new("id", DataType::Int32, false)]));
+
+        db.create_table(CreateTableRequest {
+            name: "source".to_string(),
+            namespace: vec![],
+            data: CreateTableData::Empty(TableDefinition::new_from_schema(schema)),
+            mode: CreateTableMode::Create,
+            write_options: Default::default(),
+        })
+        .await
+        .unwrap();
+
+        let source_uri = db.table_uri("source").unwrap();
+
+        // Try clone with both version and tag (should fail)
+        let result = db
+            .clone_table(CloneTableRequest {
+                target_table_name: "cloned".to_string(),
+                target_namespace: vec![],
+                source_uri,
+                source_version: Some(1),
+                source_tag: Some("v1.0".to_string()),
+                is_shallow: true,
+            })
+            .await;
+
+        assert!(result.is_err());
+        assert!(matches!(
+            result.unwrap_err(),
+            Error::InvalidInput { message } if message.contains("Cannot specify both source_version and source_tag")
+        ));
+    }
+
+    #[tokio::test]
+    async fn test_clone_table_with_specific_version() {
+        let (_tempdir, db) = setup_database().await;
+
+        // Create a source table with initial data
+        let schema = Arc::new(Schema::new(vec![
+            Field::new("id", DataType::Int32, false),
+            Field::new("value", DataType::Utf8, false),
+        ]));
+
+        let batch1 = RecordBatch::try_new(
+            schema.clone(),
+            vec![
+                Arc::new(Int32Array::from(vec![1, 2])),
+                Arc::new(StringArray::from(vec!["a", "b"])),
+            ],
+        )
+        .unwrap();
+
+        let reader = Box::new(arrow_array::RecordBatchIterator::new(
+            vec![Ok(batch1)],
+            schema.clone(),
+        ));
+
+        let source_table = db
+            .create_table(CreateTableRequest {
+                name: "versioned_source".to_string(),
+                namespace: vec![],
+                data: CreateTableData::Data(reader),
+                mode: CreateTableMode::Create,
+                write_options: Default::default(),
+            })
+            .await
+            .unwrap();
+
+        // Get the initial version
+        let initial_version = source_table.version().await.unwrap();
+
+        // Add more data to create a new version
+        let batch2 = RecordBatch::try_new(
+            schema.clone(),
+            vec![
+                Arc::new(Int32Array::from(vec![3, 4])),
+                Arc::new(StringArray::from(vec!["c", "d"])),
+            ],
+        )
+        .unwrap();
+
+        let source_table_obj = Table::new(source_table.clone());
+        source_table_obj
+            .add(Box::new(arrow_array::RecordBatchIterator::new(
+                vec![Ok(batch2)],
+                schema.clone(),
+            )))
+            .execute()
+            .await
+            .unwrap();
+
+        // Verify source table now has 4 rows
+        assert_eq!(source_table.count_rows(None).await.unwrap(), 4);
+
+        let source_uri = db.table_uri("versioned_source").unwrap();
+
+        // Clone from the initial version (should have only 2 rows)
+        let cloned_table = db
+            .clone_table(CloneTableRequest {
+                target_table_name: "cloned_from_version".to_string(),
+                target_namespace: vec![],
+                source_uri,
+                source_version: Some(initial_version),
+                source_tag: None,
+                is_shallow: true,
+            })
+            .await
+            .unwrap();
+
+        // Verify cloned table has only the initial 2 rows
+        assert_eq!(cloned_table.count_rows(None).await.unwrap(), 2);
+
+        // Source table should still have 4 rows
+        assert_eq!(source_table.count_rows(None).await.unwrap(), 4);
+    }
+
+    #[tokio::test]
+    async fn test_clone_table_with_tag() {
+        let (_tempdir, db) = setup_database().await;
+
+        // Create a source table with initial data
+        let schema = Arc::new(Schema::new(vec![
+            Field::new("id", DataType::Int32, false),
+            Field::new("value", DataType::Utf8, false),
+        ]));
+
+        let batch1 = RecordBatch::try_new(
+            schema.clone(),
+            vec![
+                Arc::new(Int32Array::from(vec![1, 2])),
+                Arc::new(StringArray::from(vec!["a", "b"])),
+            ],
+        )
+        .unwrap();
+
+        let reader = Box::new(arrow_array::RecordBatchIterator::new(
+            vec![Ok(batch1)],
+            schema.clone(),
+        ));
+
+        let source_table = db
+            .create_table(CreateTableRequest {
+                name: "tagged_source".to_string(),
+                namespace: vec![],
+                data: CreateTableData::Data(reader),
+                mode: CreateTableMode::Create,
+                write_options: Default::default(),
+            })
+            .await
+            .unwrap();
+
+        // Create a tag for the current version
+        let source_table_obj = Table::new(source_table.clone());
+        let mut tags = source_table_obj.tags().await.unwrap();
+        tags.create("v1.0", source_table.version().await.unwrap())
+            .await
+            .unwrap();
+
+        // Add more data after the tag
+        let batch2 = RecordBatch::try_new(
+            schema.clone(),
+            vec![
+                Arc::new(Int32Array::from(vec![3, 4])),
+                Arc::new(StringArray::from(vec!["c", "d"])),
+            ],
+        )
+        .unwrap();
+
+        let source_table_obj = Table::new(source_table.clone());
+        source_table_obj
+            .add(Box::new(arrow_array::RecordBatchIterator::new(
+                vec![Ok(batch2)],
+                schema.clone(),
+            )))
+            .execute()
+            .await
+            .unwrap();
+
+        // Source table should have 4 rows
+        assert_eq!(source_table.count_rows(None).await.unwrap(), 4);
+
+        let source_uri = db.table_uri("tagged_source").unwrap();
+
+        // Clone from the tag (should have only 2 rows)
+        let cloned_table = db
+            .clone_table(CloneTableRequest {
+                target_table_name: "cloned_from_tag".to_string(),
+                target_namespace: vec![],
+                source_uri,
+                source_version: None,
+                source_tag: Some("v1.0".to_string()),
+                is_shallow: true,
+            })
+            .await
+            .unwrap();
+
+        // Verify cloned table has only the tagged version's 2 rows
+        assert_eq!(cloned_table.count_rows(None).await.unwrap(), 2);
+    }
+
+    #[tokio::test]
+    async fn test_cloned_tables_evolve_independently() {
+        let (_tempdir, db) = setup_database().await;
+
+        // Create a source table with initial data
+        let schema = Arc::new(Schema::new(vec![
+            Field::new("id", DataType::Int32, false),
+            Field::new("value", DataType::Utf8, false),
+        ]));
+
+        let batch1 = RecordBatch::try_new(
+            schema.clone(),
+            vec![
+                Arc::new(Int32Array::from(vec![1, 2])),
+                Arc::new(StringArray::from(vec!["a", "b"])),
+            ],
+        )
+        .unwrap();
+
+        let reader = Box::new(arrow_array::RecordBatchIterator::new(
+            vec![Ok(batch1)],
+            schema.clone(),
+        ));
+
+        let source_table = db
+            .create_table(CreateTableRequest {
+                name: "independent_source".to_string(),
+                namespace: vec![],
+                data: CreateTableData::Data(reader),
+                mode: CreateTableMode::Create,
+                write_options: Default::default(),
+            })
+            .await
+            .unwrap();
+
+        let source_uri = db.table_uri("independent_source").unwrap();
+
+        // Clone the table
+        let cloned_table = db
+            .clone_table(CloneTableRequest {
+                target_table_name: "independent_clone".to_string(),
+                target_namespace: vec![],
+                source_uri,
+                source_version: None,
+                source_tag: None,
+                is_shallow: true,
+            })
+            .await
+            .unwrap();
+
+        // Both should start with 2 rows
+        assert_eq!(source_table.count_rows(None).await.unwrap(), 2);
+        assert_eq!(cloned_table.count_rows(None).await.unwrap(), 2);
+
+        // Add data to the cloned table
+        let batch_clone = RecordBatch::try_new(
+            schema.clone(),
+            vec![
+                Arc::new(Int32Array::from(vec![3, 4, 5])),
+                Arc::new(StringArray::from(vec!["c", "d", "e"])),
+            ],
+        )
+        .unwrap();
+
+        let cloned_table_obj = Table::new(cloned_table.clone());
+        cloned_table_obj
+            .add(Box::new(arrow_array::RecordBatchIterator::new(
+                vec![Ok(batch_clone)],
+                schema.clone(),
+            )))
+            .execute()
+            .await
+            .unwrap();
+
+        // Add different data to the source table
+        let batch_source = RecordBatch::try_new(
+            schema.clone(),
+            vec![
+                Arc::new(Int32Array::from(vec![10, 11])),
+                Arc::new(StringArray::from(vec!["x", "y"])),
+            ],
+        )
+        .unwrap();
+
+        let source_table_obj = Table::new(source_table.clone());
+        source_table_obj
+            .add(Box::new(arrow_array::RecordBatchIterator::new(
+                vec![Ok(batch_source)],
+                schema.clone(),
+            )))
+            .execute()
+            .await
+            .unwrap();
+
+        // Verify they have evolved independently
+        assert_eq!(source_table.count_rows(None).await.unwrap(), 4); // 2 + 2
+        assert_eq!(cloned_table.count_rows(None).await.unwrap(), 5); // 2 + 3
+    }
+
+    #[tokio::test]
+    async fn test_clone_latest_version() {
+        let (_tempdir, db) = setup_database().await;
+
+        // Create a source table with initial data
+        let schema = Arc::new(Schema::new(vec![Field::new("id", DataType::Int32, false)]));
+
+        let batch1 =
+            RecordBatch::try_new(schema.clone(), vec![Arc::new(Int32Array::from(vec![1, 2]))])
+                .unwrap();
+
+        let reader = Box::new(arrow_array::RecordBatchIterator::new(
+            vec![Ok(batch1)],
+            schema.clone(),
+        ));
+
+        let source_table = db
+            .create_table(CreateTableRequest {
+                name: "latest_version_source".to_string(),
+                namespace: vec![],
+                data: CreateTableData::Data(reader),
+                mode: CreateTableMode::Create,
+                write_options: Default::default(),
+            })
+            .await
+            .unwrap();
+
+        // Add more data to create new versions
+        for i in 0..3 {
+            let batch = RecordBatch::try_new(
+                schema.clone(),
+                vec![Arc::new(Int32Array::from(vec![i * 10, i * 10 + 1]))],
+            )
+            .unwrap();
+
+            let source_table_obj = Table::new(source_table.clone());
+            source_table_obj
+                .add(Box::new(arrow_array::RecordBatchIterator::new(
+                    vec![Ok(batch)],
+                    schema.clone(),
+                )))
+                .execute()
+                .await
+                .unwrap();
+        }
+
+        // Source should have 8 rows total (2 + 2 + 2 + 2)
+        let source_count = source_table.count_rows(None).await.unwrap();
+        assert_eq!(source_count, 8);
+
+        let source_uri = db.table_uri("latest_version_source").unwrap();
+
+        // Clone without specifying version or tag (should get latest)
+        let cloned_table = db
+            .clone_table(CloneTableRequest {
+                target_table_name: "cloned_latest".to_string(),
+                target_namespace: vec![],
+                source_uri,
+                source_version: None,
+                source_tag: None,
+                is_shallow: true,
+            })
+            .await
+            .unwrap();
+
+        // Cloned table should have all 8 rows from the latest version
+        assert_eq!(cloned_table.count_rows(None).await.unwrap(), 8);
+    }
+}

rust/lancedb/src/embeddings.rs

@@ -45,10 +45,10 @@ use crate::{
 pub trait EmbeddingFunction: std::fmt::Debug + Send + Sync {
     fn name(&self) -> &str;
     /// The type of the input data
-    fn source_type(&self) -> Result<Cow<DataType>>;
+    fn source_type(&self) -> Result<Cow<'_, DataType>>;
     /// The type of the output data
     /// This should **always** match the output of the `embed` function
-    fn dest_type(&self) -> Result<Cow<DataType>>;
+    fn dest_type(&self) -> Result<Cow<'_, DataType>>;
     /// Compute the embeddings for the source column in the database
     fn compute_source_embeddings(&self, source: Arc<dyn Array>) -> Result<Arc<dyn Array>>;
     /// Compute the embeddings for a given user query

rust/lancedb/src/embeddings/bedrock.rs

@@ -75,11 +75,11 @@ impl EmbeddingFunction for BedrockEmbeddingFunction {
         "bedrock"
     }
 
-    fn source_type(&self) -> Result<Cow<DataType>> {
+    fn source_type(&self) -> Result<Cow<'_, DataType>> {
         Ok(Cow::Owned(DataType::Utf8))
     }
 
-    fn dest_type(&self) -> Result<Cow<DataType>> {
+    fn dest_type(&self) -> Result<Cow<'_, DataType>> {
         let n_dims = self.model.ndims();
         Ok(Cow::Owned(DataType::new_fixed_size_list(
             DataType::Float32,

rust/lancedb/src/embeddings/openai.rs

@@ -144,11 +144,11 @@ impl EmbeddingFunction for OpenAIEmbeddingFunction {
         "openai"
     }
 
-    fn source_type(&self) -> Result<Cow<DataType>> {
+    fn source_type(&self) -> Result<Cow<'_, DataType>> {
         Ok(Cow::Owned(DataType::Utf8))
     }
 
-    fn dest_type(&self) -> Result<Cow<DataType>> {
+    fn dest_type(&self) -> Result<Cow<'_, DataType>> {
         let n_dims = self.model.ndims();
         Ok(Cow::Owned(DataType::new_fixed_size_list(
             DataType::Float32,

rust/lancedb/src/embeddings/sentence_transformers.rs

@@ -407,11 +407,11 @@ impl EmbeddingFunction for SentenceTransformersEmbeddings {
         "sentence-transformers"
     }
 
-    fn source_type(&self) -> crate::Result<std::borrow::Cow<arrow_schema::DataType>> {
+    fn source_type(&self) -> crate::Result<std::borrow::Cow<'_, arrow_schema::DataType>> {
         Ok(Cow::Owned(DataType::Utf8))
     }
 
-    fn dest_type(&self) -> crate::Result<std::borrow::Cow<arrow_schema::DataType>> {
+    fn dest_type(&self) -> crate::Result<std::borrow::Cow<'_, arrow_schema::DataType>> {
         let (n_dims, dtype) = self.compute_ndims_and_dtype()?;
         Ok(Cow::Owned(DataType::new_fixed_size_list(
             dtype,

rust/lancedb/src/index.rs

@@ -65,12 +65,94 @@ pub enum Index {
 /// Builder for the create_index operation
 ///
 /// The methods on this builder are used to specify options common to all indices.
+///
+/// # Examples
+///
+/// Creating a basic vector index:
+///
+/// ```
+/// use lancedb::{connect, index::{Index, vector::IvfPqIndexBuilder}};
+///
+/// # async fn create_basic_vector_index() -> lancedb::Result<()> {
+/// let db = connect("data/sample-lancedb").execute().await?;
+/// let table = db.open_table("my_table").execute().await?;
+///
+/// // Create a vector index with default settings
+/// table
+///     .create_index(&["vector"], Index::IvfPq(IvfPqIndexBuilder::default()))
+///     .execute()
+///     .await?;
+/// # Ok(())
+/// # }
+/// ```
+///
+/// Creating an index with a custom name:
+///
+/// ```
+/// use lancedb::{connect, index::{Index, vector::IvfPqIndexBuilder}};
+///
+/// # async fn create_named_index() -> lancedb::Result<()> {
+/// let db = connect("data/sample-lancedb").execute().await?;
+/// let table = db.open_table("my_table").execute().await?;
+///
+/// // Create a vector index with a custom name
+/// table
+///     .create_index(&["embeddings"], Index::IvfPq(IvfPqIndexBuilder::default()))
+///     .name("my_embeddings_index".to_string())
+///     .execute()
+///     .await?;
+/// # Ok(())
+/// # }
+/// ```
+///
+/// Creating an untrained index (for scalar indices only):
+///
+/// ```
+/// use lancedb::{connect, index::{Index, scalar::BTreeIndexBuilder}};
+///
+/// # async fn create_untrained_index() -> lancedb::Result<()> {
+/// let db = connect("data/sample-lancedb").execute().await?;
+/// let table = db.open_table("my_table").execute().await?;
+///
+/// // Create a BTree index without training (creates empty index)
+/// table
+///     .create_index(&["category"], Index::BTree(BTreeIndexBuilder::default()))
+///     .train(false)
+///     .name("category_index".to_string())
+///     .execute()
+///     .await?;
+/// # Ok(())
+/// # }
+/// ```
+///
+/// Creating a scalar index with all options:
+///
+/// ```
+/// use lancedb::{connect, index::{Index, scalar::BitmapIndexBuilder}};
+///
+/// # async fn create_full_options_index() -> lancedb::Result<()> {
+/// let db = connect("data/sample-lancedb").execute().await?;
+/// let table = db.open_table("my_table").execute().await?;
+///
+/// // Create a bitmap index with full configuration
+/// table
+///     .create_index(&["status"], Index::Bitmap(BitmapIndexBuilder::default()))
+///     .name("status_bitmap_index".to_string())
+///     .train(true)  // Train the index with existing data
+///     .replace(false)  // Don't replace if index already exists
+///     .execute()
+///     .await?;
+/// # Ok(())
+/// # }
+/// ```
 pub struct IndexBuilder {
     parent: Arc<dyn BaseTable>,
     pub(crate) index: Index,
     pub(crate) columns: Vec<String>,
     pub(crate) replace: bool,
     pub(crate) wait_timeout: Option<Duration>,
+    pub(crate) train: bool,
+    pub(crate) name: Option<String>,
 }
 
 impl IndexBuilder {
@@ -80,7 +162,9 @@ impl IndexBuilder {
             index,
             columns,
             replace: true,
+            train: true,
             wait_timeout: None,
+            name: None,
         }
     }
 
@@ -94,6 +178,82 @@ impl IndexBuilder {
         self
     }
 
+    /// The name of the index. If not set, a default name will be generated.
+    ///
+    /// # Examples
+    ///
+    /// ```
+    /// use lancedb::{connect, index::{Index, scalar::BTreeIndexBuilder}};
+    ///
+    /// # async fn name_example() -> lancedb::Result<()> {
+    /// let db = connect("data/sample-lancedb").execute().await?;
+    /// let table = db.open_table("my_table").execute().await?;
+    ///
+    /// // Create an index with a custom name
+    /// table
+    ///     .create_index(&["user_id"], Index::BTree(BTreeIndexBuilder::default()))
+    ///     .name("user_id_btree_index".to_string())
+    ///     .execute()
+    ///     .await?;
+    /// # Ok(())
+    /// # }
+    /// ```
+    pub fn name(mut self, v: String) -> Self {
+        self.name = Some(v);
+        self
+    }
+
+    /// Whether to train the index, the default is `true`.
+    ///
+    /// If this is false, the index will not be trained and just created empty.
+    ///
+    /// This is not supported for vector indices yet.
+    ///
+    /// # Examples
+    ///
+    /// Creating an empty index that will be populated later:
+    ///
+    /// ```
+    /// use lancedb::{connect, index::{Index, scalar::BitmapIndexBuilder}};
+    ///
+    /// # async fn train_false_example() -> lancedb::Result<()> {
+    /// let db = connect("data/sample-lancedb").execute().await?;
+    /// let table = db.open_table("my_table").execute().await?;
+    ///
+    /// // Create an empty bitmap index (not trained with existing data)
+    /// table
+    ///     .create_index(&["category"], Index::Bitmap(BitmapIndexBuilder::default()))
+    ///     .train(false)  // Create empty index
+    ///     .name("category_bitmap".to_string())
+    ///     .execute()
+    ///     .await?;
+    /// # Ok(())
+    /// # }
+    /// ```
+    ///
+    /// Creating a trained index (default behavior):
+    ///
+    /// ```
+    /// use lancedb::{connect, index::{Index, scalar::BTreeIndexBuilder}};
+    ///
+    /// # async fn train_true_example() -> lancedb::Result<()> {
+    /// let db = connect("data/sample-lancedb").execute().await?;
+    /// let table = db.open_table("my_table").execute().await?;
+    ///
+    /// // Create a trained BTree index (includes existing data)
+    /// table
+    ///     .create_index(&["timestamp"], Index::BTree(BTreeIndexBuilder::default()))
+    ///     .train(true)  // Train with existing data (this is the default)
+    ///     .execute()
+    ///     .await?;
+    /// # Ok(())
+    /// # }
+    /// ```
+    pub fn train(mut self, v: bool) -> Self {
+        self.train = v;
+        self
+    }
+
     /// Duration of time to wait for asynchronous indexing to complete. If not set,
     /// `create_index()` will not wait.
     ///

rust/lancedb/src/index/vector.rs

@@ -8,7 +8,7 @@
 //! values
 use std::cmp::max;
 
-use lance::table::format::{Index, Manifest};
+use lance::table::format::{IndexMetadata, Manifest};
 
 use crate::DistanceType;
 
@@ -19,7 +19,7 @@ pub struct VectorIndex {
 }
 
 impl VectorIndex {
-    pub fn new_from_format(manifest: &Manifest, index: &Index) -> Self {
+    pub fn new_from_format(manifest: &Manifest, index: &IndexMetadata) -> Self {
         let fields = index
             .fields
             .iter()
@@ -112,6 +112,15 @@ macro_rules! impl_ivf_params_setter {
             self.max_iterations = max_iterations;
             self
         }
+
+        /// The target size of each partition.
+        ///
+        /// This value controls the tradeoff between search performance and accuracy.
+        /// The higher the value the faster the search but the less accurate the results will be.
+        pub fn target_partition_size(mut self, target_partition_size: u32) -> Self {
+            self.target_partition_size = Some(target_partition_size);
+            self
+        }
     };
 }
 
@@ -182,6 +191,7 @@ pub struct IvfFlatIndexBuilder {
     pub(crate) num_partitions: Option<u32>,
     pub(crate) sample_rate: u32,
     pub(crate) max_iterations: u32,
+    pub(crate) target_partition_size: Option<u32>,
 }
 
 impl Default for IvfFlatIndexBuilder {
@@ -191,6 +201,7 @@ impl Default for IvfFlatIndexBuilder {
             num_partitions: None,
             sample_rate: 256,
             max_iterations: 50,
+            target_partition_size: None,
         }
     }
 }
@@ -228,6 +239,7 @@ pub struct IvfPqIndexBuilder {
     pub(crate) num_partitions: Option<u32>,
     pub(crate) sample_rate: u32,
     pub(crate) max_iterations: u32,
+    pub(crate) target_partition_size: Option<u32>,
 
     // PQ
     pub(crate) num_sub_vectors: Option<u32>,
@@ -243,6 +255,7 @@ impl Default for IvfPqIndexBuilder {
             num_bits: None,
             sample_rate: 256,
             max_iterations: 50,
+            target_partition_size: None,
         }
     }
 }
@@ -293,6 +306,7 @@ pub struct IvfHnswPqIndexBuilder {
     pub(crate) num_partitions: Option<u32>,
     pub(crate) sample_rate: u32,
     pub(crate) max_iterations: u32,
+    pub(crate) target_partition_size: Option<u32>,
 
     // HNSW
     pub(crate) m: u32,
@@ -314,6 +328,7 @@ impl Default for IvfHnswPqIndexBuilder {
             max_iterations: 50,
             m: 20,
             ef_construction: 300,
+            target_partition_size: None,
         }
     }
 }
@@ -341,6 +356,7 @@ pub struct IvfHnswSqIndexBuilder {
     pub(crate) num_partitions: Option<u32>,
     pub(crate) sample_rate: u32,
     pub(crate) max_iterations: u32,
+    pub(crate) target_partition_size: Option<u32>,
 
     // HNSW
     pub(crate) m: u32,
@@ -358,6 +374,7 @@ impl Default for IvfHnswSqIndexBuilder {
             max_iterations: 50,
             m: 20,
             ef_construction: 300,
+            target_partition_size: None,
         }
     }
 }

rust/lancedb/src/io/object_store.rs

@@ -9,7 +9,7 @@ use futures::{stream::BoxStream, TryFutureExt};
 use lance::io::WrappingObjectStore;
 use object_store::{
     path::Path, Error, GetOptions, GetResult, ListResult, MultipartUpload, ObjectMeta, ObjectStore,
-    PutMultipartOpts, PutOptions, PutPayload, PutResult, Result, UploadPart,
+    PutMultipartOptions, PutOptions, PutPayload, PutResult, Result, UploadPart,
 };
 
 use async_trait::async_trait;
@@ -73,7 +73,7 @@ impl ObjectStore for MirroringObjectStore {
     async fn put_multipart_opts(
         &self,
         location: &Path,
-        opts: PutMultipartOpts,
+        opts: PutMultipartOptions,
     ) -> Result<Box<dyn MultipartUpload>> {
         if location.primary_only() {
             return self.primary.put_multipart_opts(location, opts).await;
@@ -170,7 +170,11 @@ impl MirroringObjectStoreWrapper {
 }
 
 impl WrappingObjectStore for MirroringObjectStoreWrapper {
-    fn wrap(&self, primary: Arc<dyn ObjectStore>) -> Arc<dyn ObjectStore> {
+    fn wrap(
+        &self,
+        primary: Arc<dyn ObjectStore>,
+        _storage_options: Option<&std::collections::HashMap<String, String>>,
+    ) -> Arc<dyn ObjectStore> {
         Arc::new(MirroringObjectStore {
             primary,
             secondary: self.secondary.clone(),

rust/lancedb/src/io/object_store/io_tracking.rs

@@ -11,7 +11,7 @@ use futures::stream::BoxStream;
 use lance::io::WrappingObjectStore;
 use object_store::{
     path::Path, GetOptions, GetResult, ListResult, MultipartUpload, ObjectMeta, ObjectStore,
-    PutMultipartOpts, PutOptions, PutPayload, PutResult, Result as OSResult, UploadPart,
+    PutMultipartOptions, PutOptions, PutPayload, PutResult, Result as OSResult, UploadPart,
 };
 
 #[derive(Debug, Default)]
@@ -50,7 +50,11 @@ impl IoStatsHolder {
 }
 
 impl WrappingObjectStore for IoStatsHolder {
-    fn wrap(&self, target: Arc<dyn ObjectStore>) -> Arc<dyn ObjectStore> {
+    fn wrap(
+        &self,
+        target: Arc<dyn ObjectStore>,
+        _storage_options: Option<&std::collections::HashMap<String, String>>,
+    ) -> Arc<dyn ObjectStore> {
         Arc::new(IoTrackingStore {
             target,
             stats: self.0.clone(),
@@ -106,7 +110,7 @@ impl ObjectStore for IoTrackingStore {
     async fn put_multipart_opts(
         &self,
         location: &Path,
-        opts: PutMultipartOpts,
+        opts: PutMultipartOptions,
     ) -> OSResult<Box<dyn MultipartUpload>> {
         let target = self.target.put_multipart_opts(location, opts).await?;
         Ok(Box::new(IoTrackingMultipartUpload {