Bump version: 0.22.1-beta.0 → 0.22.1-beta.1

.bumpversion.toml

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

.github/workflows/docs_test.yml

@@ -24,8 +24,7 @@ env:
 jobs:
   test-python:
     name: Test doc python code
-    runs-on: warp-ubuntu-2204-x64-8x
+    runs-on: ubuntu-24.04
-    timeout-minutes: 60
     steps:
     - name: Checkout
       uses: actions/checkout@v4
@@ -49,6 +48,7 @@ jobs:
       uses: swatinem/rust-cache@v2
     - name: Build Python
       working-directory: docs/test
+      timeout-minutes: 60
       run:
         python -m pip install --extra-index-url https://pypi.fury.io/lancedb/ -r requirements.txt
     - name: Create test files

.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: 90
+    timeout-minutes: 60
     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: warp-macos-14-arm64-6x
+            runner: macos-14
     env:
       MACOSX_DEPLOYMENT_TARGET: 10.15
     steps:

Cargo.lock

@@ -713,9 +713,9 @@ dependencies = [
 
 [[package]]
 name = "aws-sdk-s3"
-version = "1.105.0"
+version = "1.104.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c99789e929b5e1d9a5aa3fa1d81317f3a789afc796141d11b0eaafd9d9f47e38"
+checksum = "38c488cd6abb0ec9811c401894191932e941c5f91dc226043edacd0afa1634bc"
 dependencies = [
  "aws-credential-types",
  "aws-runtime",
@@ -963,9 +963,9 @@ dependencies = [
 
 [[package]]
 name = "aws-smithy-runtime"
-version = "1.9.2"
+version = "1.9.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4fa63ad37685ceb7762fa4d73d06f1d5493feb88e3f27259b9ed277f4c01b185"
+checksum = "d3946acbe1ead1301ba6862e712c7903ca9bb230bdf1fbd1b5ac54158ef2ab1f"
 dependencies = [
  "aws-smithy-async",
  "aws-smithy-http",
@@ -2929,18 +2929,6 @@ version = "0.2.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "f8eb564c5c7423d25c886fb561d1e4ee69f72354d16918afa32c08811f6b6a55"
 
-[[package]]
-name = "fastbloom"
-version = "0.14.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "18c1ddb9231d8554c2d6bdf4cfaabf0c59251658c68b6c95cd52dd0c513a912a"
-dependencies = [
- "getrandom 0.3.3",
- "libm",
- "rand 0.9.2",
- "siphasher",
-]
-
 [[package]]
 name = "fastdivide"
 version = "0.4.2"
@@ -3040,9 +3028,8 @@ checksum = "42703706b716c37f96a77aea830392ad231f44c9e9a67872fa5548707e11b11c"
 
 [[package]]
 name = "fsst"
-version = "0.37.0"
+version = "0.35.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/lancedb/lance.git?tag=v0.35.0-beta.4#e842a8f922b90c298c356dd1c6afdc83ca5253f2"
-checksum = "fe0a0b1d16ce6b863be8ab766004d89ebf0779fd6ce31b0ef3bbc7fedaaad373"
 dependencies = [
  "arrow-array",
  "rand 0.9.2",
@@ -4219,9 +4206,8 @@ dependencies = [
 
 [[package]]
 name = "lance"
-version = "0.37.0"
+version = "0.35.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/lancedb/lance.git?tag=v0.35.0-beta.4#e842a8f922b90c298c356dd1c6afdc83ca5253f2"
-checksum = "42171f2af5d377e6bbcc8a8572144ee15b73a8f78ceb6160f1adeabf0d0f3e3c"
 dependencies = [
  "arrow",
  "arrow-arith",
@@ -4284,9 +4270,8 @@ dependencies = [
 
 [[package]]
 name = "lance-arrow"
-version = "0.37.0"
+version = "0.35.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/lancedb/lance.git?tag=v0.35.0-beta.4#e842a8f922b90c298c356dd1c6afdc83ca5253f2"
-checksum = "25ef9499a1e581112f45fbf743fdc8e24830cda0bd13396b11c71aa6e6cba083"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -4304,9 +4289,8 @@ dependencies = [
 
 [[package]]
 name = "lance-bitpacking"
-version = "0.37.0"
+version = "0.35.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/lancedb/lance.git?tag=v0.35.0-beta.4#e842a8f922b90c298c356dd1c6afdc83ca5253f2"
-checksum = "1101fffd5b161bbdc6e932d6c0a7f94cb1752b0f8cd6d18ef9064052ab901a84"
 dependencies = [
  "arrayref",
  "paste",
@@ -4315,9 +4299,8 @@ dependencies = [
 
 [[package]]
 name = "lance-core"
-version = "0.37.0"
+version = "0.35.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/lancedb/lance.git?tag=v0.35.0-beta.4#e842a8f922b90c298c356dd1c6afdc83ca5253f2"
-checksum = "527ee5e6472d058d8c66c702fbe318a3f60f971e652e60dcfc6349bdbc9b0733"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -4352,9 +4335,8 @@ dependencies = [
 
 [[package]]
 name = "lance-datafusion"
-version = "0.37.0"
+version = "0.35.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/lancedb/lance.git?tag=v0.35.0-beta.4#e842a8f922b90c298c356dd1c6afdc83ca5253f2"
-checksum = "65a80f7f15f2d941ec7b8253625cbb8e12081ea27584dd1fbc657fb9fb377f7a"
 dependencies = [
  "arrow",
  "arrow-array",
@@ -4383,9 +4365,8 @@ dependencies = [
 
 [[package]]
 name = "lance-datagen"
-version = "0.37.0"
+version = "0.35.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/lancedb/lance.git?tag=v0.35.0-beta.4#e842a8f922b90c298c356dd1c6afdc83ca5253f2"
-checksum = "0495c8afa18f246ac4b337c47d7827560283783963dd2177862d91161478fd79"
 dependencies = [
  "arrow",
  "arrow-array",
@@ -4402,9 +4383,8 @@ dependencies = [
 
 [[package]]
 name = "lance-encoding"
-version = "0.37.0"
+version = "0.35.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/lancedb/lance.git?tag=v0.35.0-beta.4#e842a8f922b90c298c356dd1c6afdc83ca5253f2"
-checksum = "0e80e9ae49d68b95d58e77d9177f68983dce4f0803ef42840e1631b38dd66adc"
 dependencies = [
  "arrow-arith",
  "arrow-array",
@@ -4432,7 +4412,6 @@ dependencies = [
  "prost-types",
  "rand 0.9.2",
  "snafu",
- "strum",
  "tokio",
  "tracing",
  "xxhash-rust",
@@ -4441,9 +4420,8 @@ dependencies = [
 
 [[package]]
 name = "lance-file"
-version = "0.37.0"
+version = "0.35.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/lancedb/lance.git?tag=v0.35.0-beta.4#e842a8f922b90c298c356dd1c6afdc83ca5253f2"
-checksum = "f1707f9f5097b36c82d3a8524bb41c762c80d5dfa5e32aa7bfc6a1c0847a1cce"
 dependencies = [
  "arrow-arith",
  "arrow-array",
@@ -4477,9 +4455,8 @@ dependencies = [
 
 [[package]]
 name = "lance-index"
-version = "0.37.0"
+version = "0.35.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/lancedb/lance.git?tag=v0.35.0-beta.4#e842a8f922b90c298c356dd1c6afdc83ca5253f2"
-checksum = "28ab52586a5a7f5371a5abf4862968231f8c0232ce0780bc456f1ec16e9370f9"
 dependencies = [
  "arrow",
  "arrow-array",
@@ -4500,7 +4477,6 @@ dependencies = [
  "datafusion-sql",
  "deepsize",
  "dirs",
- "fastbloom",
  "fst",
  "futures",
  "half",
@@ -4515,7 +4491,6 @@ dependencies = [
  "lance-io",
  "lance-linalg",
  "lance-table",
- "libm",
  "log",
  "num-traits",
  "object_store",
@@ -4532,15 +4507,13 @@ dependencies = [
  "tempfile",
  "tokio",
  "tracing",
- "twox-hash",
  "uuid",
 ]
 
 [[package]]
 name = "lance-io"
-version = "0.37.0"
+version = "0.35.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/lancedb/lance.git?tag=v0.35.0-beta.4#e842a8f922b90c298c356dd1c6afdc83ca5253f2"
-checksum = "d606f9f6a7f8ec2cacf28dfce7b2fc39e7db9f0ec77f907b8e47c756e3dd163b"
 dependencies = [
  "arrow",
  "arrow-arith",
@@ -4580,9 +4553,8 @@ dependencies = [
 
 [[package]]
 name = "lance-linalg"
-version = "0.37.0"
+version = "0.35.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/lancedb/lance.git?tag=v0.35.0-beta.4#e842a8f922b90c298c356dd1c6afdc83ca5253f2"
-checksum = "c9f1a94a5d966ff1eae817a835e3a57b34f73300f83a43bb28e7e2806695b8ba"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -4605,9 +4577,8 @@ dependencies = [
 
 [[package]]
 name = "lance-table"
-version = "0.37.0"
+version = "0.35.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/lancedb/lance.git?tag=v0.35.0-beta.4#e842a8f922b90c298c356dd1c6afdc83ca5253f2"
-checksum = "fac5c0ca6e5c285645465b95fb99fc464a1fd22a6d4b32ae0e0760f06b4b8a7f"
 dependencies = [
  "arrow",
  "arrow-array",
@@ -4645,9 +4616,8 @@ dependencies = [
 
 [[package]]
 name = "lance-testing"
-version = "0.37.0"
+version = "0.35.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/lancedb/lance.git?tag=v0.35.0-beta.4#e842a8f922b90c298c356dd1c6afdc83ca5253f2"
-checksum = "384acc1dd13379a2ae24f3e3635d9c1f4fb4dc1534f7ffd2740c268f2eb73455"
 dependencies = [
  "arrow-array",
  "arrow-schema",
@@ -4658,7 +4628,7 @@ dependencies = [
 
 [[package]]
 name = "lancedb"
-version = "0.22.1-beta.3"
+version = "0.22.1-beta.0"
 dependencies = [
  "arrow",
  "arrow-array",
@@ -4745,7 +4715,7 @@ dependencies = [
 
 [[package]]
 name = "lancedb-nodejs"
-version = "0.22.1-beta.3"
+version = "0.22.1-beta.0"
 dependencies = [
  "arrow-array",
  "arrow-ipc",
@@ -4765,7 +4735,7 @@ dependencies = [
 
 [[package]]
 name = "lancedb-python"
-version = "0.25.1-beta.3"
+version = "0.25.1-beta.0"
 dependencies = [
  "arrow",
  "async-trait",
@@ -7811,15 +7781,6 @@ version = "0.11.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "7da8b5736845d9f2fcb837ea5d9e2628564b3b043a70948a3f0b778838c5fb4f"
 
-[[package]]
-name = "strum"
-version = "0.25.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "290d54ea6f91c969195bdbcd7442c8c2a2ba87da8bf60a7ee86a235d4bc1e125"
-dependencies = [
- "strum_macros",
-]
-
 [[package]]
 name = "strum_macros"
 version = "0.25.3"
@@ -8480,9 +8441,6 @@ name = "twox-hash"
 version = "2.1.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "8b907da542cbced5261bd3256de1b3a1bf340a3d37f93425a07362a1d687de56"
-dependencies = [
- "rand 0.9.2",
-]
 
 [[package]]
 name = "typenum"

Cargo.toml

@@ -15,14 +15,14 @@ categories = ["database-implementations"]
 rust-version = "1.78.0"
 
 [workspace.dependencies]
-lance = { "version" = "=0.37.0", default-features = false, "features" = ["dynamodb"] }
+lance = { "version" = "=0.35.0", default-features = false, "features" = ["dynamodb"], "tag" = "v0.35.0-beta.4", "git" = "https://github.com/lancedb/lance.git" }
-lance-io = { "version" = "=0.37.0", default-features = false }
+lance-io = { "version" = "=0.35.0", default-features = false, "tag" = "v0.35.0-beta.4", "git" = "https://github.com/lancedb/lance.git" }
-lance-index = "=0.37.0"
+lance-index = { "version" = "=0.35.0", "tag" = "v0.35.0-beta.4", "git" = "https://github.com/lancedb/lance.git" }
-lance-linalg = "=0.37.0"
+lance-linalg = { "version" = "=0.35.0", "tag" = "v0.35.0-beta.4", "git" = "https://github.com/lancedb/lance.git" }
-lance-table = "=0.37.0"
+lance-table = { "version" = "=0.35.0", "tag" = "v0.35.0-beta.4", "git" = "https://github.com/lancedb/lance.git" }
-lance-testing = "=0.37.0"
+lance-testing = { "version" = "=0.35.0", "tag" = "v0.35.0-beta.4", "git" = "https://github.com/lancedb/lance.git" }
-lance-datafusion = "=0.37.0"
+lance-datafusion = { "version" = "=0.35.0", "tag" = "v0.35.0-beta.4", "git" = "https://github.com/lancedb/lance.git" }
-lance-encoding = "=0.37.0"
+lance-encoding = { "version" = "=0.35.0", "tag" = "v0.35.0-beta.4", "git" = "https://github.com/lancedb/lance.git" }
 # Note that this one does not include pyarrow
 arrow = { version = "55.1", optional = false }
 arrow-array = "55.1"

ci/set_lance_version.py

@@ -1,5 +1,4 @@
 import argparse
-import re
 import sys
 import json
 
@@ -19,12 +18,8 @@ def run_command(command: str) -> str:
 
 def get_latest_stable_version() -> str:
     version_line = run_command("cargo info lance | grep '^version:'")
-    # Example output: "version: 0.35.0 (latest 0.37.0)"
+    version = version_line.split(" ")[1].strip()
-    match = re.search(r'\(latest ([0-9.]+)\)', version_line)
+    return version
-    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:

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

@@ -16,7 +16,6 @@ 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

@@ -6,7 +6,6 @@ use jni::JNIEnv;
 
 use crate::Result;
 
-#[allow(dead_code)]
 pub trait FromJObject<T> {
     fn extract(&self) -> Result<T>;
 }
@@ -40,7 +39,6 @@ impl FromJObject<f64> for JObject<'_> {
     }
 }
 
-#[allow(dead_code)]
 pub trait FromJString {
     fn extract(&self, env: &mut JNIEnv) -> Result<String>;
 }
@@ -68,7 +66,6 @@ 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>0.22.1-beta.3</version>
+        <version>0.22.1-beta.1</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>0.22.1-beta.3</version>
+        <version>0.22.1-beta.1</version>
         <relativePath>../pom.xml</relativePath>
     </parent>
 

java/pom.xml

@@ -6,7 +6,7 @@
 
     <groupId>com.lancedb</groupId>
     <artifactId>lancedb-parent</artifactId>
-    <version>0.22.1-beta.3</version>
+    <version>0.22.1-beta.1</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 = "0.22.1-beta.3"
+version = "0.22.1-beta.1"
 license.workspace = true
 description.workspace = true
 repository.workspace = true

nodejs/__test__/arrow.test.ts

@@ -1008,64 +1008,5 @@ 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

@@ -203,106 +203,3 @@ 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

@@ -256,60 +256,6 @@ 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/lancedb/arrow.ts

@@ -512,11 +512,7 @@ function* rowPathsAndValues(
     if (isObject(value)) {
       yield* rowPathsAndValues(value, [...basePath, key]);
     } else {
-      // Skip undefined values - they should be treated the same as missing fields
+      yield [[...basePath, key], value];
-      // for embedding function purposes
-      if (value !== undefined) {
-        yield [[...basePath, key], value];
-      }
     }
   }
 }
@@ -705,7 +701,7 @@ function transposeData(
       }
       return current;
     });
-    return makeVector(values, field.type, undefined, field.nullable);
+    return makeVector(values, field.type);
   }
 }
 
@@ -752,30 +748,9 @@ 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) {
@@ -900,12 +875,7 @@ 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(
+      columns[field.name] = makeVector(nullValues, field.type);
-        nullValues,
-        field.type,
-        undefined,
-        field.nullable,
-      );
     }
   }
 
@@ -969,12 +939,7 @@ async function applyEmbeddings<T>(
     } else if (schema != null) {
       const destField = schema.fields.find((f) => f.name === destColumn);
       if (destField != null) {
-        newColumns[destColumn] = makeVector(
+        newColumns[destColumn] = makeVector([], destField.type);
-          [],
-          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

@@ -268,33 +268,6 @@ export abstract class Connection {
    * @param {string[]} namespace The namespace to drop tables from (defaults to root namespace).
    */
   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 */
@@ -359,28 +332,6 @@ 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 {

nodejs/npm/darwin-arm64/package.json

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

nodejs/npm/darwin-x64/package.json

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

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

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

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

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

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

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

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

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

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

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

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

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

nodejs/package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "@lancedb/lancedb",
-  "version": "0.22.1-beta.3",
+  "version": "0.22.1-beta.0",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "@lancedb/lancedb",
-      "version": "0.22.1-beta.3",
+      "version": "0.22.1-beta.0",
       "cpu": [
         "x64",
         "arm64"

nodejs/package.json

@@ -11,7 +11,7 @@
     "ann"
   ],
   "private": false,
-  "version": "0.22.1-beta.3",
+  "version": "0.22.1-beta.1",
   "main": "dist/index.js",
   "exports": {
     ".": "./dist/index.js",

nodejs/src/connection.rs

@@ -213,36 +213,6 @@ 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, namespace: Vec<String>) -> napi::Result<()> {

python/.bumpversion.toml

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

python/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "lancedb-python"
-version = "0.25.1"
+version = "0.25.1-beta.1"
 edition.workspace = true
 description = "Python bindings for LanceDB"
 license.workspace = true

python/python/lancedb/_lancedb.pyi

@@ -60,15 +60,6 @@ class Connection(object):
         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,

python/python/lancedb/db.py

@@ -665,60 +665,6 @@ class LanceDBConnection(DBConnection):
             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,
@@ -1190,54 +1136,6 @@ class AsyncConnection(object):
         )
         return AsyncTable(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,
-    ) -> 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,

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"]).create(**obj["model"]),
+                function=self.get(obj["name"])(**obj["model"]),
             )
             for obj in raw_list
         }

python/python/lancedb/index.py

@@ -251,13 +251,6 @@ 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"
@@ -268,7 +261,6 @@ class HnswPq:
     sample_rate: int = 256
     m: int = 20
     ef_construction: int = 300
-    target_partition_size: Optional[int] = None
 
 
 @dataclass
@@ -359,12 +351,6 @@ 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"
@@ -373,7 +359,6 @@ class HnswSq:
     sample_rate: int = 256
     m: int = 20
     ef_construction: int = 300
-    target_partition_size: Optional[int] = None
 
 
 @dataclass
@@ -459,20 +444,12 @@ 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
@@ -587,13 +564,6 @@ 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"
@@ -602,7 +572,6 @@ class IvfPq:
     num_bits: int = 8
     max_iterations: int = 50
     sample_rate: int = 256
-    target_partition_size: Optional[int] = None
 
 
 __all__ = [

python/python/lancedb/remote/db.py

@@ -212,53 +212,6 @@ class RemoteDBConnection(DBConnection):
         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
     def create_table(
         self,

python/python/lancedb/rerankers/__init__.py

@@ -9,7 +9,6 @@ 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
 
@@ -24,5 +23,4 @@ __all__ = [
     "RRFReranker",
     "AnswerdotaiRerankers",
     "VoyageAIReranker",
-    "MRRReranker",
 ]

python/python/lancedb/rerankers/mrr.py

@@ -1,169 +0,0 @@
-# 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

@@ -691,7 +691,6 @@ class Table(ABC):
         ef_construction: int = 300,
         name: Optional[str] = None,
         train: bool = True,
-        target_partition_size: Optional[int] = None,
     ):
         """Create an index on the table.
 
@@ -1470,7 +1469,10 @@ 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
-            This parameter is no longer used and is deprecated.
+            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.
 
         Experimental API
         ----------------
@@ -2000,7 +2002,6 @@ class LanceTable(Table):
         *,
         name: Optional[str] = None,
         train: bool = True,
-        target_partition_size: Optional[int] = None,
     ):
         """Create an index on the table."""
         if accelerator is not None:
@@ -2017,7 +2018,6 @@ class LanceTable(Table):
                 num_bits=num_bits,
                 m=m,
                 ef_construction=ef_construction,
-                target_partition_size=target_partition_size,
             )
             self.checkout_latest()
             return
@@ -2027,7 +2027,6 @@ 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(
@@ -2037,7 +2036,6 @@ 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(
@@ -2049,7 +2047,6 @@ 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(
@@ -2059,7 +2056,6 @@ 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}")
@@ -2832,7 +2828,10 @@ 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
-            This parameter is no longer used and is deprecated.
+            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.
 
         Experimental API
         ----------------
@@ -4292,7 +4291,10 @@ 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
-            This parameter is no longer used and is deprecated.
+            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.
 
         Experimental API
         ----------------
@@ -4315,19 +4317,10 @@ 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

@@ -747,16 +747,15 @@ def test_local_namespace_operations(tmp_path):
     # Create a local database connection
     db = lancedb.connect(tmp_path)
 
-    # Test list_namespaces returns empty list for root namespace
+    # Test list_namespaces returns empty list
     namespaces = list(db.list_namespaces())
     assert namespaces == []
 
-    # Test list_namespaces with non-empty namespace raises NotImplementedError
+    # Test list_namespaces with parameters still returns empty list
-    with pytest.raises(
+    namespaces_with_params = list(
-        NotImplementedError,
+        db.list_namespaces(namespace=["test"], page_token="token", limit=5)
-        match="Namespace operations are not supported for listing database",
+    )
-    ):
+    assert namespaces_with_params == []
-        list(db.list_namespaces(namespace=["test"]))
 
 
 def test_local_create_namespace_not_supported(tmp_path):
@@ -831,119 +830,3 @@ def test_local_table_operations_with_namespace_raise_error(tmp_path):
     # 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

@@ -114,63 +114,6 @@ 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_rerankers.py

@@ -22,7 +22,6 @@ from lancedb.rerankers import (
     JinaReranker,
     AnswerdotaiRerankers,
     VoyageAIReranker,
-    MRRReranker,
 )
 from lancedb.table import LanceTable
 
@@ -47,7 +46,6 @@ 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
@@ -98,7 +96,7 @@ def get_test_table(tmp_path, use_tantivy):
     )
 
     # Create a fts index
-    table.create_fts_index("text", use_tantivy=use_tantivy, replace=True)
+    table.create_fts_index("text", use_tantivy=use_tantivy)
 
     return table, MyTable
 
@@ -322,34 +320,6 @@ 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

@@ -674,45 +674,6 @@ def test_create_index_method(mock_create_index, mem_db: DBConnection):
         "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",
         metric="dot",

python/src/connection.rs

@@ -163,34 +163,6 @@ impl Connection {
         })
     }
 
-    #[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>,
@@ -283,7 +255,7 @@ 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>,

python/src/index.rs

@@ -63,9 +63,6 @@ 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" => {
@@ -79,9 +76,6 @@ 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);
                 }
@@ -100,9 +94,6 @@ 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);
                 }
@@ -120,9 +111,6 @@ 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!(
@@ -156,7 +144,6 @@ struct IvfFlatParams {
     num_partitions: Option<u32>,
     max_iterations: u32,
     sample_rate: u32,
-    target_partition_size: Option<u32>,
 }
 
 #[derive(FromPyObject)]
@@ -167,7 +154,6 @@ struct IvfPqParams {
     num_bits: u32,
     max_iterations: u32,
     sample_rate: u32,
-    target_partition_size: Option<u32>,
 }
 
 #[derive(FromPyObject)]
@@ -180,7 +166,6 @@ struct IvfHnswPqParams {
     sample_rate: u32,
     m: u32,
     ef_construction: u32,
-    target_partition_size: Option<u32>,
 }
 
 #[derive(FromPyObject)]
@@ -191,7 +176,6 @@ struct IvfHnswSqParams {
     sample_rate: u32,
     m: u32,
     ef_construction: u32,
-    target_partition_size: Option<u32>,
 }
 
 #[pyclass(get_all)]

python/src/table.rs

@@ -591,11 +591,12 @@ 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))]
+    #[pyo3(signature = (cleanup_since_ms=None, delete_unverified=None, retrain=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 {
@@ -631,9 +632,10 @@ impl Table {
                 .prune
                 .unwrap();
             inner
-                .optimize(lancedb::table::OptimizeAction::Index(
+                .optimize(lancedb::table::OptimizeAction::Index(match retrain {
-                    OptimizeOptions::default(),
+                    Some(true) => OptimizeOptions::retrain(),
-                ))
+                    _ => OptimizeOptions::default(),
+                }))
                 .await
                 .infer_error()?;
             Ok(OptimizeStats {

rust/lancedb/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "lancedb"
-version = "0.22.1-beta.3"
+version = "0.22.1-beta.1"
 edition.workspace = true
 description = "LanceDB: A serverless, low-latency vector database for AI applications"
 license.workspace = true
@@ -86,11 +86,11 @@ 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 = "1.55.0" }
+aws-sdk-dynamodb = { version = "1.38.0" }
-aws-sdk-s3 = { version = "1.55.0" }
+aws-sdk-s3 = { version = "1.38.0" }
-aws-sdk-kms = { version = "1.48.0" }
+aws-sdk-kms = { version = "1.37" }
-aws-config = { version = "1.5.10" }
+aws-config = { version = "1.0" }
-aws-smithy-runtime = { version = "1.9.1" }
+aws-smithy-runtime = { version = "1.3" }
 datafusion.workspace = true
 http-body = "1"                                        # Matching reqwest
 rstest = "0.23.0"

rust/lancedb/src/catalog.rs

@@ -0,0 +1,86 @@
+// 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

@@ -0,0 +1,624 @@
+// 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(),
+            namespace: vec![],
+        })
+        .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(),
+            namespace: vec![],
+        })
+        .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

@@ -13,13 +13,15 @@ use lance::dataset::ReadParams;
 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::{
-    CloneTableRequest, CreateNamespaceRequest, CreateTableData, CreateTableMode,
+    CreateNamespaceRequest, CreateTableData, CreateTableMode, CreateTableRequest, Database,
-    CreateTableRequest, Database, DatabaseOptions, DropNamespaceRequest, ListNamespacesRequest,
+    DatabaseOptions, DropNamespaceRequest, ListNamespacesRequest, OpenTableRequest,
-    OpenTableRequest, TableNamesRequest,
+    TableNamesRequest,
 };
 use crate::embeddings::{
     EmbeddingDefinition, EmbeddingFunction, EmbeddingRegistry, MemoryRegistry, WithEmbeddings,
@@ -469,62 +471,6 @@ 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 {
@@ -631,30 +577,6 @@ 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.
@@ -738,7 +660,7 @@ pub struct ConnectRequest {
     #[cfg(feature = "remote")]
     pub client_config: ClientConfig,
 
-    /// Database specific options
+    /// Database/Catalog specific options
     pub options: HashMap<String, String>,
 
     /// The interval at which to check for updates from other processes.
@@ -1015,6 +937,50 @@ 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::*;
@@ -1056,6 +1022,7 @@ 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};
@@ -1363,48 +1330,89 @@ mod tests {
     }
 
     #[tokio::test]
-    async fn test_clone_table() {
+    async fn test_connect_catalog() {
         let tmp_dir = tempdir().unwrap();
         let uri = tmp_dir.path().to_str().unwrap();
-        let db = connect(uri).execute().await.unwrap();
+        let catalog = connect_catalog(uri).execute().await.unwrap();
 
-        // Create a source table with some data
+        // Verify that we can get the uri from the catalog
-        let mut batch_gen = BatchGenerator::new()
+        let catalog_uri = catalog.uri();
-            .col(Box::new(IncrementingInt32::new().named("id")))
+        assert_eq!(catalog_uri, uri);
-            .col(Box::new(IncrementingInt32::new().named("value")));
-        let reader = batch_gen.batches(5, 100);
 
-        let source_table = db
+        // Check that the catalog is initially empty
-            .create_table("source_table", reader)
+        let dbs = catalog
-            .execute()
+            .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(),
+            })
             .await
             .unwrap();
 
-        // Get the source table URI
+        let dbs = catalog
-        let source_table_path = tmp_dir.path().join("source_table.lance");
+            .database_names(DatabaseNamesRequest::default())
-        let source_uri = source_table_path.to_str().unwrap();
+            .await
+            .unwrap();
+        assert_eq!(dbs.len(), 1);
+        assert_eq!(dbs[0], db_name);
 
-        // Clone the table
+        let db = catalog
-        let cloned_table = db
+            .open_database(OpenDatabaseRequest {
-            .clone_table("cloned_table", source_uri)
+                name: db_name.to_string(),
-            .execute()
+                database_options: HashMap::new(),
+            })
             .await
             .unwrap();
 
-        // Verify the cloned table exists
+        let tables = db.table_names(Default::default()).await.unwrap();
-        let table_names = db.table_names().execute().await.unwrap();
+        assert_eq!(tables.len(), 0);
-        assert!(table_names.contains(&"source_table".to_string()));
+    }
-        assert!(table_names.contains(&"cloned_table".to_string()));
 
-        // Verify the cloned table has the same schema
+    #[tokio::test]
-        assert_eq!(
+    #[cfg(not(windows))]
-            source_table.schema().await.unwrap(),
+    async fn test_catalog_drop_database() {
-            cloned_table.schema().await.unwrap()
+        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 data
+        // Create and then drop a database
-        let source_count = source_table.count_rows(None).await.unwrap();
+        let db_name = "test_db_to_drop";
-        let cloned_count = cloned_table.count_rows(None).await.unwrap();
+        catalog
-        assert_eq!(source_count, cloned_count);
+            .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);
     }
 }

rust/lancedb/src/database.rs

@@ -176,42 +176,6 @@ 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.
@@ -229,13 +193,6 @@ pub trait 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

rust/lancedb/src/database/listing.rs

@@ -7,8 +7,7 @@ use std::fs::create_dir_all;
 use std::path::Path;
 use std::{collections::HashMap, sync::Arc};
 
-use lance::dataset::refs::Ref;
+use lance::dataset::{ReadParams, WriteMode};
-use lance::dataset::{builder::DatasetBuilder, ReadParams, WriteMode};
 use lance::io::{ObjectStore, ObjectStoreParams, WrappingObjectStore};
 use lance_datafusion::utils::StreamingWriteSource;
 use lance_encoding::version::LanceFileVersion;
@@ -23,8 +22,8 @@ use crate::table::NativeTable;
 use crate::utils::validate_table_name;
 
 use super::{
-    BaseTable, CloneTableRequest, CreateNamespaceRequest, CreateTableMode, CreateTableRequest,
+    BaseTable, CreateNamespaceRequest, CreateTableMode, CreateTableRequest, Database,
-    Database, DatabaseOptions, DropNamespaceRequest, ListNamespacesRequest, OpenTableRequest,
+    DatabaseOptions, DropNamespaceRequest, ListNamespacesRequest, OpenTableRequest,
     TableNamesRequest,
 };
 
@@ -588,13 +587,7 @@ impl ListingDatabase {
 
 #[async_trait::async_trait]
 impl Database for ListingDatabase {
-    async fn list_namespaces(&self, request: ListNamespacesRequest) -> Result<Vec<String>> {
+    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())
     }
 
@@ -685,65 +678,6 @@ impl Database for ListingDatabase {
         }
     }
 
-    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 {
@@ -845,694 +779,3 @@ 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/vector.rs

@@ -8,7 +8,7 @@
 //! values
 use std::cmp::max;
 
-use lance::table::format::{IndexMetadata, Manifest};
+use lance::table::format::{Index, Manifest};
 
 use crate::DistanceType;
 
@@ -19,7 +19,7 @@ pub struct VectorIndex {
 }
 
 impl VectorIndex {
-    pub fn new_from_format(manifest: &Manifest, index: &IndexMetadata) -> Self {
+    pub fn new_from_format(manifest: &Manifest, index: &Index) -> Self {
         let fields = index
             .fields
             .iter()
@@ -112,15 +112,6 @@ 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
-        }
     };
 }
 
@@ -191,7 +182,6 @@ 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 {
@@ -201,7 +191,6 @@ impl Default for IvfFlatIndexBuilder {
             num_partitions: None,
             sample_rate: 256,
             max_iterations: 50,
-            target_partition_size: None,
         }
     }
 }
@@ -239,7 +228,6 @@ 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>,
@@ -255,7 +243,6 @@ impl Default for IvfPqIndexBuilder {
             num_bits: None,
             sample_rate: 256,
             max_iterations: 50,
-            target_partition_size: None,
         }
     }
 }
@@ -306,7 +293,6 @@ 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,
@@ -328,7 +314,6 @@ impl Default for IvfHnswPqIndexBuilder {
             max_iterations: 50,
             m: 20,
             ef_construction: 300,
-            target_partition_size: None,
         }
     }
 }
@@ -356,7 +341,6 @@ 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,
@@ -374,7 +358,6 @@ impl Default for IvfHnswSqIndexBuilder {
             max_iterations: 50,
             m: 20,
             ef_construction: 300,
-            target_partition_size: None,
         }
     }
 }

rust/lancedb/src/lib.rs

@@ -191,6 +191,7 @@
 //! ```
 
 pub mod arrow;
+pub mod catalog;
 pub mod connection;
 pub mod data;
 pub mod database;

rust/lancedb/src/remote/db.rs

@@ -14,9 +14,9 @@ use serde::Deserialize;
 use tokio::task::spawn_blocking;
 
 use crate::database::{
-    CloneTableRequest, CreateNamespaceRequest, CreateTableData, CreateTableMode,
+    CreateNamespaceRequest, CreateTableData, CreateTableMode, CreateTableRequest, Database,
-    CreateTableRequest, Database, DatabaseOptions, DropNamespaceRequest, ListNamespacesRequest,
+    DatabaseOptions, DropNamespaceRequest, ListNamespacesRequest, OpenTableRequest,
-    OpenTableRequest, TableNamesRequest,
+    TableNamesRequest,
 };
 use crate::error::Result;
 use crate::table::BaseTable;
@@ -27,18 +27,6 @@ use super::table::RemoteTable;
 use super::util::{batches_to_ipc_bytes, parse_server_version};
 use super::ARROW_STREAM_CONTENT_TYPE;
 
-// Request structure for the remote clone table API
-#[derive(serde::Serialize)]
-struct RemoteCloneTableRequest {
-    source_location: String,
-    #[serde(skip_serializing_if = "Option::is_none")]
-    source_version: Option<u64>,
-    #[serde(skip_serializing_if = "Option::is_none")]
-    source_tag: Option<String>,
-    #[serde(skip_serializing_if = "Option::is_none")]
-    is_shallow: Option<bool>,
-}
-
 // the versions of the server that we support
 // for any new feature that we need to change the SDK behavior, we should bump the server version,
 // and add a feature flag as method of `ServerVersion` here.
@@ -442,51 +430,6 @@ impl<S: HttpSend> Database for RemoteDatabase<S> {
         Ok(table)
     }
 
-    async fn clone_table(&self, request: CloneTableRequest) -> Result<Arc<dyn BaseTable>> {
-        let table_identifier = build_table_identifier(
-            &request.target_table_name,
-            &request.target_namespace,
-            &self.client.id_delimiter,
-        );
-
-        let remote_request = RemoteCloneTableRequest {
-            source_location: request.source_uri,
-            source_version: request.source_version,
-            source_tag: request.source_tag,
-            is_shallow: Some(request.is_shallow),
-        };
-
-        let req = self
-            .client
-            .post(&format!("/v1/table/{}/clone", table_identifier.clone()))
-            .json(&remote_request);
-
-        let (request_id, rsp) = self.client.send(req).await?;
-
-        let status = rsp.status();
-        if status != StatusCode::OK {
-            let body = rsp.text().await.err_to_http(request_id.clone())?;
-            return Err(crate::Error::Http {
-                source: format!("Failed to clone table: {}", body).into(),
-                request_id,
-                status_code: Some(status),
-            });
-        }
-
-        let version = parse_server_version(&request_id, &rsp)?;
-        let cache_key = build_cache_key(&request.target_table_name, &request.target_namespace);
-        let table = Arc::new(RemoteTable::new(
-            self.client.clone(),
-            request.target_table_name.clone(),
-            request.target_namespace.clone(),
-            table_identifier,
-            version,
-        ));
-        self.table_cache.insert(cache_key, table.clone()).await;
-
-        Ok(table)
-    }
-
     async fn open_table(&self, request: OpenTableRequest) -> Result<Arc<dyn BaseTable>> {
         let identifier =
             build_table_identifier(&request.name, &request.namespace, &self.client.id_delimiter);
@@ -1278,146 +1221,4 @@ mod tests {
             _ => panic!("Expected Runtime error from header provider"),
         }
     }
-
-    #[tokio::test]
-    async fn test_clone_table() {
-        let conn = Connection::new_with_handler(|request| {
-            assert_eq!(request.method(), &reqwest::Method::POST);
-            assert_eq!(request.url().path(), "/v1/table/cloned_table/clone");
-            assert_eq!(
-                request.headers().get("Content-Type").unwrap(),
-                JSON_CONTENT_TYPE
-            );
-
-            let body = request.body().unwrap().as_bytes().unwrap();
-            let body: serde_json::Value = serde_json::from_slice(body).unwrap();
-            assert_eq!(body["source_location"], "s3://bucket/source_table");
-            assert_eq!(body["is_shallow"], true);
-
-            http::Response::builder().status(200).body("").unwrap()
-        });
-
-        let table = conn
-            .clone_table("cloned_table", "s3://bucket/source_table")
-            .execute()
-            .await
-            .unwrap();
-        assert_eq!(table.name(), "cloned_table");
-    }
-
-    #[tokio::test]
-    async fn test_clone_table_with_version() {
-        let conn = Connection::new_with_handler(|request| {
-            assert_eq!(request.method(), &reqwest::Method::POST);
-            assert_eq!(request.url().path(), "/v1/table/cloned_table/clone");
-
-            let body = request.body().unwrap().as_bytes().unwrap();
-            let body: serde_json::Value = serde_json::from_slice(body).unwrap();
-            assert_eq!(body["source_location"], "s3://bucket/source_table");
-            assert_eq!(body["source_version"], 42);
-            assert_eq!(body["is_shallow"], true);
-
-            http::Response::builder().status(200).body("").unwrap()
-        });
-
-        let table = conn
-            .clone_table("cloned_table", "s3://bucket/source_table")
-            .source_version(42)
-            .execute()
-            .await
-            .unwrap();
-        assert_eq!(table.name(), "cloned_table");
-    }
-
-    #[tokio::test]
-    async fn test_clone_table_with_tag() {
-        let conn = Connection::new_with_handler(|request| {
-            assert_eq!(request.method(), &reqwest::Method::POST);
-            assert_eq!(request.url().path(), "/v1/table/cloned_table/clone");
-
-            let body = request.body().unwrap().as_bytes().unwrap();
-            let body: serde_json::Value = serde_json::from_slice(body).unwrap();
-            assert_eq!(body["source_location"], "s3://bucket/source_table");
-            assert_eq!(body["source_tag"], "v1.0");
-            assert_eq!(body["is_shallow"], true);
-
-            http::Response::builder().status(200).body("").unwrap()
-        });
-
-        let table = conn
-            .clone_table("cloned_table", "s3://bucket/source_table")
-            .source_tag("v1.0")
-            .execute()
-            .await
-            .unwrap();
-        assert_eq!(table.name(), "cloned_table");
-    }
-
-    #[tokio::test]
-    async fn test_clone_table_deep_clone() {
-        let conn = Connection::new_with_handler(|request| {
-            assert_eq!(request.method(), &reqwest::Method::POST);
-            assert_eq!(request.url().path(), "/v1/table/cloned_table/clone");
-
-            let body = request.body().unwrap().as_bytes().unwrap();
-            let body: serde_json::Value = serde_json::from_slice(body).unwrap();
-            assert_eq!(body["source_location"], "s3://bucket/source_table");
-            assert_eq!(body["is_shallow"], false);
-
-            http::Response::builder().status(200).body("").unwrap()
-        });
-
-        let table = conn
-            .clone_table("cloned_table", "s3://bucket/source_table")
-            .is_shallow(false)
-            .execute()
-            .await
-            .unwrap();
-        assert_eq!(table.name(), "cloned_table");
-    }
-
-    #[tokio::test]
-    async fn test_clone_table_with_namespace() {
-        let conn = Connection::new_with_handler(|request| {
-            assert_eq!(request.method(), &reqwest::Method::POST);
-            assert_eq!(request.url().path(), "/v1/table/ns1$ns2$cloned_table/clone");
-
-            let body = request.body().unwrap().as_bytes().unwrap();
-            let body: serde_json::Value = serde_json::from_slice(body).unwrap();
-            assert_eq!(body["source_location"], "s3://bucket/source_table");
-            assert_eq!(body["is_shallow"], true);
-
-            http::Response::builder().status(200).body("").unwrap()
-        });
-
-        let table = conn
-            .clone_table("cloned_table", "s3://bucket/source_table")
-            .target_namespace(vec!["ns1".to_string(), "ns2".to_string()])
-            .execute()
-            .await
-            .unwrap();
-        assert_eq!(table.name(), "cloned_table");
-    }
-
-    #[tokio::test]
-    async fn test_clone_table_error() {
-        let conn = Connection::new_with_handler(|_| {
-            http::Response::builder()
-                .status(500)
-                .body("Internal server error")
-                .unwrap()
-        });
-
-        let result = conn
-            .clone_table("cloned_table", "s3://bucket/source_table")
-            .execute()
-            .await;
-
-        assert!(result.is_err());
-        if let Err(crate::Error::Http { source, .. }) = result {
-            assert!(source.to_string().contains("Failed to clone table"));
-        } else {
-            panic!("Expected HTTP error");
-        }
-    }
 }

rust/lancedb/src/table.rs

@@ -242,15 +242,17 @@ pub struct OptimizeStats {
 /// Describes what happens when a vector either contains NaN or
 /// does not have enough values
 #[derive(Clone, Debug, Default)]
-#[allow(dead_code)] // https://github.com/lancedb/lancedb/issues/992
 enum BadVectorHandling {
     /// An error is returned
     #[default]
     Error,
+    #[allow(dead_code)] // https://github.com/lancedb/lancedb/issues/992
     /// The offending row is droppped
     Drop,
+    #[allow(dead_code)] // https://github.com/lancedb/lancedb/issues/992
     /// The invalid/missing items are replaced by fill_value
     Fill(f32),
+    #[allow(dead_code)] // https://github.com/lancedb/lancedb/issues/992
     /// The invalid items are replaced by NULL
     None,
 }
@@ -1976,8 +1978,6 @@ impl NativeTable {
     /// Delete keys from the config
     pub async fn delete_config_keys(&self, delete_keys: &[&str]) -> Result<()> {
         let mut dataset = self.dataset.get_mut().await?;
-        // TODO: update this when we implement metadata APIs
-        #[allow(deprecated)]
         dataset.delete_config_keys(delete_keys).await?;
         Ok(())
     }
@@ -1988,8 +1988,6 @@ impl NativeTable {
         upsert_values: impl IntoIterator<Item = (String, String)>,
     ) -> Result<()> {
         let mut dataset = self.dataset.get_mut().await?;
-        // TODO: update this when we implement metadata APIs
-        #[allow(deprecated)]
         dataset.replace_schema_metadata(upsert_values).await?;
         Ok(())
     }

rust/lancedb/src/utils.rs

@@ -20,8 +20,6 @@ use datafusion_physical_plan::SendableRecordBatchStream;
 
 lazy_static! {
     static ref TABLE_NAME_REGEX: regex::Regex = regex::Regex::new(r"^[a-zA-Z0-9_\-\.]+$").unwrap();
-    static ref NAMESPACE_NAME_REGEX: regex::Regex =
-        regex::Regex::new(r"^[a-zA-Z0-9_\-\.]+$").unwrap();
 }
 
 pub trait PatchStoreParam {
@@ -100,53 +98,6 @@ pub fn validate_table_name(name: &str) -> Result<()> {
     Ok(())
 }
 
-/// Validate a namespace name component
-///
-/// Namespace names must:
-/// - Not be empty
-/// - Only contain alphanumeric characters, underscores, hyphens, and periods
-///
-/// # Arguments
-/// * `name` - A single namespace component (not the full path)
-///
-/// # Returns
-/// * `Ok(())` if the namespace name is valid
-/// * `Err(Error)` if the namespace name is invalid
-pub fn validate_namespace_name(name: &str) -> Result<()> {
-    if name.is_empty() {
-        return Err(Error::InvalidInput {
-            message: "Namespace names cannot be empty strings".to_string(),
-        });
-    }
-    if !NAMESPACE_NAME_REGEX.is_match(name) {
-        return Err(Error::InvalidInput {
-            message: format!(
-                "Invalid namespace name '{}': Namespace names can only contain alphanumeric characters, underscores, hyphens, and periods",
-                name
-            ),
-        });
-    }
-    Ok(())
-}
-
-/// Validate all components of a namespace
-///
-/// Iterates through all namespace components and validates each one.
-/// Returns an error if any component is invalid.
-///
-/// # Arguments
-/// * `namespace` - The namespace components to validate
-///
-/// # Returns
-/// * `Ok(())` if all namespace components are valid
-/// * `Err(Error)` if any component is invalid
-pub fn validate_namespace(namespace: &[String]) -> Result<()> {
-    for component in namespace {
-        validate_namespace_name(component)?;
-    }
-    Ok(())
-}
-
 /// Find one default column to create index or perform vector query.
 pub(crate) fn default_vector_column(schema: &Schema, dim: Option<i32>) -> Result<String> {
     // Try to find a vector column.
@@ -394,61 +345,6 @@ mod tests {
         assert!(validate_table_name("name with space").is_err());
     }
 
-    #[test]
-    fn test_validate_namespace_name() {
-        // Valid namespace names
-        assert!(validate_namespace_name("ns1").is_ok());
-        assert!(validate_namespace_name("namespace_123").is_ok());
-        assert!(validate_namespace_name("my-namespace").is_ok());
-        assert!(validate_namespace_name("my.namespace").is_ok());
-        assert!(validate_namespace_name("NS_1.2.3").is_ok());
-        assert!(validate_namespace_name("a").is_ok());
-        assert!(validate_namespace_name("123").is_ok());
-        assert!(validate_namespace_name("_underscore").is_ok());
-        assert!(validate_namespace_name("-hyphen").is_ok());
-        assert!(validate_namespace_name(".period").is_ok());
-
-        // Invalid namespace names
-        assert!(validate_namespace_name("").is_err());
-        assert!(validate_namespace_name("namespace with spaces").is_err());
-        assert!(validate_namespace_name("namespace/with/slashes").is_err());
-        assert!(validate_namespace_name("namespace\\with\\backslashes").is_err());
-        assert!(validate_namespace_name("namespace$with$delimiter").is_err());
-        assert!(validate_namespace_name("namespace@special").is_err());
-        assert!(validate_namespace_name("namespace#hash").is_err());
-    }
-
-    #[test]
-    fn test_validate_namespace() {
-        // Valid namespace with single component
-        assert!(validate_namespace(&["ns1".to_string()]).is_ok());
-
-        // Valid namespace with multiple components
-        assert!(
-            validate_namespace(&["ns1".to_string(), "ns2".to_string(), "ns3".to_string()]).is_ok()
-        );
-
-        // Empty namespace (root) is valid
-        assert!(validate_namespace(&[]).is_ok());
-
-        // Invalid: contains empty component
-        assert!(validate_namespace(&["ns1".to_string(), "".to_string()]).is_err());
-
-        // Invalid: contains component with spaces
-        assert!(validate_namespace(&["ns1".to_string(), "ns 2".to_string()]).is_err());
-
-        // Invalid: contains component with special characters
-        assert!(validate_namespace(&["ns1".to_string(), "ns@2".to_string()]).is_err());
-        assert!(validate_namespace(&["ns1".to_string(), "ns/2".to_string()]).is_err());
-        assert!(validate_namespace(&["ns1".to_string(), "ns$2".to_string()]).is_err());
-
-        // Valid: underscores, hyphens, and periods are allowed
-        assert!(
-            validate_namespace(&["ns_1".to_string(), "ns-2".to_string(), "ns.3".to_string()])
-                .is_ok()
-        );
-    }
-
     #[test]
     fn test_string_to_datatype() {
         let string = "int32";

rust/lancedb/tests/embedding_registry_test.rs

@@ -341,10 +341,10 @@ impl EmbeddingFunction for MockEmbed {
     fn name(&self) -> &str {
         &self.name
     }
-    fn source_type(&self) -> Result<Cow<'_, DataType>> {
+    fn source_type(&self) -> Result<Cow<DataType>> {
         Ok(Cow::Borrowed(&self.source_type))
     }
-    fn dest_type(&self) -> Result<Cow<'_, DataType>> {
+    fn dest_type(&self) -> Result<Cow<DataType>> {
         Ok(Cow::Borrowed(&self.dest_type))
     }
     fn compute_source_embeddings(&self, source: Arc<dyn Array>) -> Result<Arc<dyn Array>> {