feat: connection pool for sync client

BREAKING-CHANGE: change Pydantic Vector field to be nullable by default.
Closes #1577

.bumpversion.toml

@@ -1,5 +1,5 @@
 [tool.bumpversion]
-current_version = "0.13.0"
+current_version = "0.13.1-beta.0"
 parse = """(?x)
     (?P<major>0|[1-9]\\d*)\\.
     (?P<minor>0|[1-9]\\d*)\\.

java/core/pom.xml

@@ -8,7 +8,7 @@
     <parent>
         <groupId>com.lancedb</groupId>
         <artifactId>lancedb-parent</artifactId>
-        <version>0.13.0-final.0</version>
+        <version>0.13.1-beta.0</version>
         <relativePath>../pom.xml</relativePath>
     </parent>
 

java/pom.xml

@@ -6,7 +6,7 @@
 
     <groupId>com.lancedb</groupId>
     <artifactId>lancedb-parent</artifactId>
-    <version>0.13.0-final.0</version>
+    <version>0.13.1-beta.0</version>
     <packaging>pom</packaging>
 
     <name>LanceDB Parent</name>

node/package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "vectordb",
-  "version": "0.13.0",
+  "version": "0.13.1-beta.0",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "vectordb",
-      "version": "0.13.0",
+      "version": "0.13.1-beta.0",
       "cpu": [
         "x64",
         "arm64"
@@ -52,12 +52,14 @@
         "uuid": "^9.0.0"
       },
       "optionalDependencies": {
-        "@lancedb/vectordb-darwin-arm64": "0.13.0",
+        "@lancedb/vectordb-darwin-arm64": "0.13.1-beta.0",
-        "@lancedb/vectordb-darwin-x64": "0.13.0",
+        "@lancedb/vectordb-darwin-x64": "0.13.1-beta.0",
-        "@lancedb/vectordb-linux-arm64-gnu": "0.13.0",
+        "@lancedb/vectordb-linux-arm64-gnu": "0.13.1-beta.0",
-        "@lancedb/vectordb-linux-x64-gnu": "0.13.0",
+        "@lancedb/vectordb-linux-arm64-musl": "0.13.1-beta.0",
-        "@lancedb/vectordb-win32-arm64-msvc": "0.13.0",
+        "@lancedb/vectordb-linux-x64-gnu": "0.13.1-beta.0",
-        "@lancedb/vectordb-win32-x64-msvc": "0.13.0"
+        "@lancedb/vectordb-linux-x64-musl": "0.13.1-beta.0",
+        "@lancedb/vectordb-win32-arm64-msvc": "0.13.1-beta.0",
+        "@lancedb/vectordb-win32-x64-msvc": "0.13.1-beta.0"
       },
       "peerDependencies": {
         "@apache-arrow/ts": "^14.0.2",
@@ -327,66 +329,6 @@
         "@jridgewell/sourcemap-codec": "^1.4.10"
       }
     },
-    "node_modules/@lancedb/vectordb-darwin-arm64": {
-      "version": "0.13.0",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-arm64/-/vectordb-darwin-arm64-0.13.0.tgz",
-      "integrity": "sha512-8hdcjkRmgrdQYf1jN+DyZae40LIv8UUfnWy70Uid5qy63sSvRW/+MvIdqIPFr9QlLUXmpyyQuX0y3bZhUR99cQ==",
-      "cpu": [
-        "arm64"
-      ],
-      "optional": true,
-      "os": [
-        "darwin"
-      ]
-    },
-    "node_modules/@lancedb/vectordb-darwin-x64": {
-      "version": "0.13.0",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-x64/-/vectordb-darwin-x64-0.13.0.tgz",
-      "integrity": "sha512-fWzAY4l5SQtNfMYh80v+M66ugZHhdxbkpk5mNEv6Zsug3DL6kRj3Uv31/i0wgzY6F5G3LUlbjZerN+eTnDLwOw==",
-      "cpu": [
-        "x64"
-      ],
-      "optional": true,
-      "os": [
-        "darwin"
-      ]
-    },
-    "node_modules/@lancedb/vectordb-linux-arm64-gnu": {
-      "version": "0.13.0",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-arm64-gnu/-/vectordb-linux-arm64-gnu-0.13.0.tgz",
-      "integrity": "sha512-ltwAT9baOSuR5YiGykQXPC8/HGYF13vpI47qxhP9yfgiz9pA8EUn8p8YrBRzq7J4DIZ4b8JSVDXQnMIqEtB4Kg==",
-      "cpu": [
-        "arm64"
-      ],
-      "optional": true,
-      "os": [
-        "linux"
-      ]
-    },
-    "node_modules/@lancedb/vectordb-linux-x64-gnu": {
-      "version": "0.13.0",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-x64-gnu/-/vectordb-linux-x64-gnu-0.13.0.tgz",
-      "integrity": "sha512-MiT/RBlMPGGRh7BX+MXwRuNiiUnKmuDcHH8nm88IH28T7TQxXIbA9w6UpSg5m9f3DgKQI2K8oLi29oKIB8ZwDQ==",
-      "cpu": [
-        "x64"
-      ],
-      "optional": true,
-      "os": [
-        "linux"
-      ]
-    },
-    "node_modules/@lancedb/vectordb-win32-x64-msvc": {
-      "version": "0.13.0",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-win32-x64-msvc/-/vectordb-win32-x64-msvc-0.13.0.tgz",
-      "integrity": "sha512-SovP/hwWYLJIy65DKbVuXlBPTb/nwvVpTO6dh9zRch+L5ek6JmVAkwsfeTS2p5bMa8VPujsCXYUAVuCDEJU8wg==",
-      "cpu": [
-        "x64"
-      ],
-      "optional": true,
-      "os": [
-        "win32"
-      ]
-    },
     "node_modules/@neon-rs/cli": {
       "version": "0.0.160",
       "resolved": "https://registry.npmjs.org/@neon-rs/cli/-/cli-0.0.160.tgz",

node/package.json

@@ -1,6 +1,6 @@
 {
   "name": "vectordb",
-  "version": "0.13.0",
+  "version": "0.13.1-beta.0",
   "description": " Serverless, low-latency vector database for AI applications",
   "main": "dist/index.js",
   "types": "dist/index.d.ts",
@@ -84,18 +84,20 @@
       "aarch64-apple-darwin": "@lancedb/vectordb-darwin-arm64",
       "x86_64-unknown-linux-gnu": "@lancedb/vectordb-linux-x64-gnu",
       "aarch64-unknown-linux-gnu": "@lancedb/vectordb-linux-arm64-gnu",
+      "x86_64-unknown-linux-musl": "@lancedb/vectordb-linux-x64-musl",
+      "aarch64-unknown-linux-musl": "@lancedb/vectordb-linux-arm64-musl",
       "x86_64-pc-windows-msvc": "@lancedb/vectordb-win32-x64-msvc",
       "aarch64-pc-windows-msvc": "@lancedb/vectordb-win32-arm64-msvc"
     }
   },
   "optionalDependencies": {
-    "@lancedb/vectordb-darwin-x64": "0.13.0",
+    "@lancedb/vectordb-darwin-x64": "0.13.1-beta.0",
-    "@lancedb/vectordb-darwin-arm64": "0.13.0",
+    "@lancedb/vectordb-darwin-arm64": "0.13.1-beta.0",
-    "@lancedb/vectordb-linux-x64-gnu": "0.13.0",
+    "@lancedb/vectordb-linux-x64-gnu": "0.13.1-beta.0",
-    "@lancedb/vectordb-linux-arm64-gnu": "0.13.0",
+    "@lancedb/vectordb-linux-arm64-gnu": "0.13.1-beta.0",
-    "@lancedb/vectordb-linux-x64-musl": "0.13.0",
+    "@lancedb/vectordb-linux-x64-musl": "0.13.1-beta.0",
-    "@lancedb/vectordb-linux-arm64-musl": "0.13.0",
+    "@lancedb/vectordb-linux-arm64-musl": "0.13.1-beta.0",
-    "@lancedb/vectordb-win32-x64-msvc": "0.13.0",
+    "@lancedb/vectordb-win32-x64-msvc": "0.13.1-beta.0",
-    "@lancedb/vectordb-win32-arm64-msvc": "0.13.0"
+    "@lancedb/vectordb-win32-arm64-msvc": "0.13.1-beta.0"
   }
 }

nodejs/Cargo.toml

@@ -1,7 +1,7 @@
 [package]
 name = "lancedb-nodejs"
 edition.workspace = true
-version = "0.13.0"
+version = "0.13.1-beta.0"
 license.workspace = true
 description.workspace = true
 repository.workspace = true

nodejs/npm/darwin-arm64/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-darwin-arm64",
-	"version": "0.13.0",
+	"version": "0.13.1-beta.0",
 	"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": "0.13.0",
+	"version": "0.13.1-beta.0",
 	"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": "0.13.0",
+	"version": "0.13.1-beta.0",
 	"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": "0.13.0",
+	"version": "0.13.1-beta.0",
 	"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": "0.13.0",
+	"version": "0.13.1-beta.0",
 	"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": "0.13.0",
+	"version": "0.13.1-beta.0",
 	"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": "0.13.0",
+  "version": "0.13.1-beta.0",
   "os": [
     "win32"
   ],

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

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-win32-x64-msvc",
-	"version": "0.13.0",
+	"version": "0.13.1-beta.0",
 	"os": ["win32"],
 	"cpu": ["x64"],
 	"main": "lancedb.win32-x64-msvc.node",

nodejs/package.json

@@ -10,7 +10,7 @@
     "vector database",
     "ann"
   ],
-  "version": "0.13.0",
+  "version": "0.13.1-beta.0",
   "main": "dist/index.js",
   "exports": {
     ".": "./dist/index.js",

python/Cargo.toml

@@ -17,11 +17,17 @@ crate-type = ["cdylib"]
 arrow = { version = "52.1", features = ["pyarrow"] }
 lancedb = { path = "../rust/lancedb", default-features = false }
 env_logger.workspace = true
-pyo3 = { version = "0.21", features = ["extension-module", "abi3-py38", "gil-refs"] }
+pyo3 = { version = "0.21", features = [
+    "extension-module",
+    "abi3-py39",
+    "gil-refs"
+] }
 # Using this fork for now: https://github.com/awestlake87/pyo3-asyncio/issues/119
 # pyo3-asyncio = { version = "0.20", features = ["attributes", "tokio-runtime"] }
-pyo3-asyncio-0-21 = { version = "0.21.0", features = ["attributes", "tokio-runtime"] }
+pyo3-asyncio-0-21 = { version = "0.21.0", features = [
-
+    "attributes",
+    "tokio-runtime"
+] }
 pin-project = "1.1.5"
 futures.workspace = true
 tokio = { version = "1.36.0", features = ["sync"] }
@@ -29,14 +35,13 @@ tokio = { version = "1.36.0", features = ["sync"] }
 [build-dependencies]
 pyo3-build-config = { version = "0.20.3", features = [
     "extension-module",
-    "abi3-py38",
+    "abi3-py39",
 ] }
 
 [features]
 default = ["default-tls", "remote"]
 fp16kernels = ["lancedb/fp16kernels"]
 remote = ["lancedb/remote"]
-
 # TLS
 default-tls = ["lancedb/default-tls"]
 native-tls = ["lancedb/native-tls"]

python/pyproject.toml

@@ -31,7 +31,6 @@ classifiers = [
     "Programming Language :: Python",
     "Programming Language :: Python :: 3",
     "Programming Language :: Python :: 3 :: Only",
-    "Programming Language :: Python :: 3.8",
     "Programming Language :: Python :: 3.9",
     "Programming Language :: Python :: 3.10",
     "Programming Language :: Python :: 3.11",

python/python/lancedb/embeddings/openai.py

@@ -83,25 +83,33 @@ class OpenAIEmbeddings(TextEmbeddingFunction):
         """
         openai = attempt_import_or_raise("openai")
 
+        valid_texts = []
+        valid_indices = []
+        for idx, text in enumerate(texts):
+            if text:
+                valid_texts.append(text)
+                valid_indices.append(idx)
+
         # TODO retry, rate limit, token limit
         try:
-            if self.name == "text-embedding-ada-002":
+            kwargs = {
-                rs = self._openai_client.embeddings.create(input=texts, model=self.name)
+                "input": valid_texts,
-            else:
+                "model": self.name,
-                kwargs = {
+            }
-                    "input": texts,
+            if self.name != "text-embedding-ada-002":
-                    "model": self.name,
+                kwargs["dimensions"] = self.dim
-                }
+
-                if self.dim:
+            rs = self._openai_client.embeddings.create(**kwargs)
-                    kwargs["dimensions"] = self.dim
+            valid_embeddings = {
-                rs = self._openai_client.embeddings.create(**kwargs)
+                idx: v.embedding for v, idx in zip(rs.data, valid_indices)
+            }
         except openai.BadRequestError:
             logging.exception("Bad request: %s", texts)
             return [None] * len(texts)
         except Exception:
             logging.exception("OpenAI embeddings error")
             raise
-        return [v.embedding for v in rs.data]
+        return [valid_embeddings.get(idx, None) for idx in range(len(texts))]
 
     @cached_property
     def _openai_client(self):

python/python/lancedb/pydantic.py

@@ -1,15 +1,5 @@
-#  Copyright 2023 LanceDB Developers
+# SPDX-License-Identifier: Apache-2.0
-#
+# SPDX-FileCopyrightText: Copyright The LanceDB Authors
-#  Licensed under the Apache License, Version 2.0 (the "License");
-#  you may not use this file except in compliance with the License.
-#  You may obtain a copy of the License at
-#      http://www.apache.org/licenses/LICENSE-2.0
-#
-#  Unless required by applicable law or agreed to in writing, software
-#  distributed under the License is distributed on an "AS IS" BASIS,
-#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-#  See the License for the specific language governing permissions and
-#  limitations under the License.
 
 """Pydantic (v1 / v2) adapter for LanceDB"""
 
@@ -30,6 +20,7 @@ from typing import (
     Type,
     Union,
     _GenericAlias,
+    GenericAlias,
 )
 
 import numpy as np
@@ -75,7 +66,7 @@ def vector(dim: int, value_type: pa.DataType = pa.float32()):
 
 
 def Vector(
-    dim: int, value_type: pa.DataType = pa.float32()
+    dim: int, value_type: pa.DataType = pa.float32(), nullable: bool = True
 ) -> Type[FixedSizeListMixin]:
     """Pydantic Vector Type.
 
@@ -88,6 +79,8 @@ def Vector(
         The dimension of the vector.
     value_type : pyarrow.DataType, optional
         The value type of the vector, by default pa.float32()
+    nullable : bool, optional
+        Whether the vector is nullable, by default it is True.
 
     Examples
     --------
@@ -103,7 +96,7 @@ def Vector(
     >>> assert schema == pa.schema([
     ...     pa.field("id", pa.int64(), False),
     ...     pa.field("url", pa.utf8(), False),
-    ...     pa.field("embeddings", pa.list_(pa.float32(), 768), False)
+    ...     pa.field("embeddings", pa.list_(pa.float32(), 768))
     ... ])
     """
 
@@ -112,6 +105,10 @@ def Vector(
         def __repr__(self):
             return f"FixedSizeList(dim={dim})"
 
+        @staticmethod
+        def nullable() -> bool:
+            return nullable
+
         @staticmethod
         def dim() -> int:
             return dim
@@ -205,9 +202,7 @@ else:
 def _pydantic_to_arrow_type(field: FieldInfo) -> pa.DataType:
     """Convert a Pydantic FieldInfo to Arrow DataType"""
 
-    if isinstance(field.annotation, _GenericAlias) or (
+    if isinstance(field.annotation, (_GenericAlias, GenericAlias)):
-        sys.version_info > (3, 9) and isinstance(field.annotation, types.GenericAlias)
-    ):
         origin = field.annotation.__origin__
         args = field.annotation.__args__
         if origin is list:
@@ -235,7 +230,7 @@ def _pydantic_to_arrow_type(field: FieldInfo) -> pa.DataType:
 
 def is_nullable(field: FieldInfo) -> bool:
     """Check if a Pydantic FieldInfo is nullable."""
-    if isinstance(field.annotation, _GenericAlias):
+    if isinstance(field.annotation, (_GenericAlias, GenericAlias)):
         origin = field.annotation.__origin__
         args = field.annotation.__args__
         if origin == Union:
@@ -246,6 +241,10 @@ def is_nullable(field: FieldInfo) -> bool:
         for typ in args:
             if typ is type(None):
                 return True
+    elif inspect.isclass(field.annotation) and issubclass(
+        field.annotation, FixedSizeListMixin
+    ):
+        return field.annotation.nullable()
     return False
 
 

python/python/lancedb/util.py

@@ -11,6 +11,8 @@ from datetime import date, datetime
 from functools import singledispatch
 from typing import Tuple, Union, Optional, Any
 from urllib.parse import urlparse
+from threading import Lock
+from contextlib import contextmanager
 
 import numpy as np
 import pyarrow as pa
@@ -314,3 +316,27 @@ def deprecated(func):
 def validate_table_name(name: str):
     """Verify the table name is valid."""
     native_validate_table_name(name)
+
+
+class ConnectionPool:
+    def __init__(self, connection_factory, *, max_size: Optional[int] = None):
+        self.max_size = max_size
+        self._connection_factory = connection_factory
+        self._pool = []
+        self._lock = Lock()
+
+    @contextmanager
+    def connection(self):
+        with self._lock:
+            if self._pool:
+                conn = self._pool.pop()
+            else:
+                conn = self._connection_factory()
+
+        # release the lock before yielding
+        try:
+            yield conn
+        finally:
+            with self._lock:
+                if self.max_size is None or len(self._pool) < self.max_size:
+                    self._pool.append(conn)

python/python/tests/test_embeddings.py

@@ -90,10 +90,13 @@ def test_embedding_with_bad_results(tmp_path):
             self, texts: Union[List[str], np.ndarray]
         ) -> list[Union[np.array, None]]:
             # Return None, which is bad if field is non-nullable
-            return [
+            a = [
-                None if i % 2 == 0 else np.random.randn(self.ndims())
+                np.full(self.ndims(), np.nan)
+                if i % 2 == 0
+                else np.random.randn(self.ndims())
                 for i in range(len(texts))
             ]
+            return a
 
     db = lancedb.connect(tmp_path)
     registry = EmbeddingFunctionRegistry.get_instance()

python/python/tests/test_embeddings_slow.py

@@ -1,15 +1,6 @@
-#  Copyright (c) 2023. LanceDB Developers
+# SPDX-License-Identifier: Apache-2.0
-#
+# SPDX-FileCopyrightText: Copyright The LanceDB Authors
-#  Licensed under the Apache License, Version 2.0 (the "License");
+
-#  you may not use this file except in compliance with the License.
-#  You may obtain a copy of the License at
-#      http://www.apache.org/licenses/LICENSE-2.0
-#
-#  Unless required by applicable law or agreed to in writing, software
-#  distributed under the License is distributed on an "AS IS" BASIS,
-#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-#  See the License for the specific language governing permissions and
-#  limitations under the License.
 import importlib
 import io
 import os
@@ -17,6 +8,7 @@ import os
 import lancedb
 import numpy as np
 import pandas as pd
+import pyarrow as pa
 import pytest
 from lancedb.embeddings import get_registry
 from lancedb.pydantic import LanceModel, Vector
@@ -444,6 +436,30 @@ def test_watsonx_embedding(tmp_path):
         assert tbl.search("hello").limit(1).to_pandas()["text"][0] == "hello world"
 
 
+@pytest.mark.slow
+@pytest.mark.skipif(
+    os.environ.get("OPENAI_API_KEY") is None, reason="OPENAI_API_KEY not set"
+)
+def test_openai_with_empty_strs(tmp_path):
+    model = get_registry().get("openai").create(max_retries=0)
+
+    class TextModel(LanceModel):
+        text: str = model.SourceField()
+        vector: Vector(model.ndims()) = model.VectorField()
+
+    df = pd.DataFrame({"text": ["hello world", ""]})
+    db = lancedb.connect(tmp_path)
+    tbl = db.create_table("test", schema=TextModel, mode="overwrite")
+
+    tbl.add(df, on_bad_vectors="skip")
+    tb = tbl.to_arrow()
+    assert tb.schema.field_by_name("vector").type == pa.list_(
+        pa.float32(), model.ndims()
+    )
+    assert len(tb) == 2
+    assert tb["vector"].is_null().to_pylist() == [False, True]
+
+
 @pytest.mark.slow
 @pytest.mark.skipif(
     importlib.util.find_spec("ollama") is None, reason="Ollama not installed"

python/python/tests/test_pydantic.py

@@ -1,16 +1,5 @@
-#  Copyright 2023 LanceDB Developers
+# SPDX-License-Identifier: Apache-2.0
-#
+# SPDX-FileCopyrightText: Copyright The LanceDB Authors
-#  Licensed under the Apache License, Version 2.0 (the "License");
-#  you may not use this file except in compliance with the License.
-#  You may obtain a copy of the License at
-#      http://www.apache.org/licenses/LICENSE-2.0
-#
-#  Unless required by applicable law or agreed to in writing, software
-#  distributed under the License is distributed on an "AS IS" BASIS,
-#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-#  See the License for the specific language governing permissions and
-#  limitations under the License.
-
 
 import json
 import sys
@@ -172,6 +161,26 @@ def test_pydantic_to_arrow_py38():
     assert schema == expect_schema
 
 
+def test_nullable_vector():
+    class NullableModel(pydantic.BaseModel):
+        vec: Vector(16, nullable=False)
+
+    schema = pydantic_to_schema(NullableModel)
+    assert schema == pa.schema([pa.field("vec", pa.list_(pa.float32(), 16), False)])
+
+    class DefaultModel(pydantic.BaseModel):
+        vec: Vector(16)
+
+    schema = pydantic_to_schema(DefaultModel)
+    assert schema == pa.schema([pa.field("vec", pa.list_(pa.float32(), 16), True)])
+
+    class NotNullableModel(pydantic.BaseModel):
+        vec: Vector(16)
+
+    schema = pydantic_to_schema(NotNullableModel)
+    assert schema == pa.schema([pa.field("vec", pa.list_(pa.float32(), 16), True)])
+
+
 def test_fixed_size_list_field():
     class TestModel(pydantic.BaseModel):
         vec: Vector(16)
@@ -192,7 +201,7 @@ def test_fixed_size_list_field():
     schema = pydantic_to_schema(TestModel)
     assert schema == pa.schema(
         [
-            pa.field("vec", pa.list_(pa.float32(), 16), False),
+            pa.field("vec", pa.list_(pa.float32(), 16)),
             pa.field("li", pa.list_(pa.int64()), False),
         ]
     )

python/python/tests/test_remote_db.py

@@ -6,13 +6,16 @@ from datetime import timedelta
 import http.server
 import json
 import threading
+from concurrent.futures import ThreadPoolExecutor
 from unittest.mock import MagicMock
 import uuid
 
 import lancedb
 from lancedb.conftest import MockTextEmbeddingFunction
 from lancedb.remote import ClientConfig
+from lancedb.util import ConnectionPool
 from lancedb.remote.errors import HttpError, RetryError
+import lancedb.util
 import pytest
 import pyarrow as pa
 
@@ -55,6 +58,34 @@ def mock_lancedb_connection(handler):
             handle.join()
 
 
+@contextlib.contextmanager
+def mock_lancedb_connection_pool(handler):
+    with http.server.HTTPServer(
+        ("localhost", 8080), make_mock_http_handler(handler)
+    ) as server:
+        handle = threading.Thread(target=server.serve_forever)
+        handle.start()
+
+        def conn_factory():
+            lancedb.connect(
+                "db://dev",
+                api_key="fake",
+                host_override="http://localhost:8080",
+                client_config={
+                    "retry_config": {"retries": 2},
+                    "timeout_config": {
+                        "connect_timeout": 1,
+                    },
+                },
+            )
+
+        try:
+            yield ConnectionPool(conn_factory)
+        finally:
+            server.shutdown()
+            handle.join()
+
+
 @contextlib.asynccontextmanager
 async def mock_lancedb_connection_async(handler):
     with http.server.HTTPServer(
@@ -187,8 +218,7 @@ async def test_retry_error():
         assert cause.status_code == 429
 
 
-@contextlib.contextmanager
+def http_handler(query_handler):
-def query_test_table(query_handler):
     def handler(request):
         if request.path == "/v1/table/test/describe/":
             request.send_response(200)
@@ -212,7 +242,12 @@ def query_test_table(query_handler):
             request.send_response(404)
             request.end_headers()
 
-    with mock_lancedb_connection(handler) as db:
+    return handler
+
+
+@contextlib.contextmanager
+def query_test_table(connection_ctx_mgr):
+    with connection_ctx_mgr as db:
         assert repr(db) == "RemoteConnect(name=dev)"
         table = db.open_table("test")
         assert repr(table) == "RemoteTable(dev.test)"
@@ -220,6 +255,7 @@ def query_test_table(query_handler):
 
 
 def test_query_sync_minimal():
+    @http_handler
     def handler(body):
         assert body == {
             "distance_type": "l2",
@@ -234,13 +270,53 @@ def test_query_sync_minimal():
 
         return pa.table({"id": [1, 2, 3]})
 
-    with query_test_table(handler) as table:
+    with query_test_table(mock_lancedb_connection(handler)) as table:
+        data = table.search([1, 2, 3]).to_list()
+        expected = [{"id": 1}, {"id": 2}, {"id": 3}]
+        assert data == expected
+
+    with query_test_table(mock_lancedb_connection_pool(handler).connection()) as table:
         data = table.search([1, 2, 3]).to_list()
         expected = [{"id": 1}, {"id": 2}, {"id": 3}]
         assert data == expected
 
 
+def test_query_sync_minimal_threaded():
+    num_query = 0
+
+    @http_handler
+    def handler(body):
+        assert body == {
+            "distance_type": "l2",
+            "k": 10,
+            "prefilter": False,
+            "refine_factor": None,
+            "ef": None,
+            "vector": [1.0, 2.0, 3.0],
+            "nprobes": 20,
+            "version": None,
+        }
+        nonlocal num_query
+        num_query += 1
+
+        return pa.table({"id": [1, 2, 3]})
+
+    pool = mock_lancedb_connection_pool(handler)
+
+    def _query(i):
+        with query_test_table(pool.connection()) as table:
+            data = table.search([1, 2, 3]).to_list()
+            expected = [{"id": 1}, {"id": 2}, {"id": 3}]
+            assert data == expected
+
+    with ThreadPoolExecutor as exec:
+        exec.map(_query, range(1000))
+
+    assert num_query == 1000
+
+
 def test_query_sync_empty_query():
+    @http_handler
     def handler(body):
         assert body == {
             "k": 10,
@@ -252,7 +328,12 @@ def test_query_sync_empty_query():
 
         return pa.table({"id": [1, 2, 3]})
 
-    with query_test_table(handler) as table:
+    with query_test_table(mock_lancedb_connection(handler)) as table:
+        data = table.search(None).where("true").select(["id"]).limit(10).to_list()
+        expected = [{"id": 1}, {"id": 2}, {"id": 3}]
+        assert data == expected
+
+    with query_test_table(mock_lancedb_connection_pool(handler).connection()) as table:
         data = table.search(None).where("true").select(["id"]).limit(10).to_list()
         expected = [{"id": 1}, {"id": 2}, {"id": 3}]
         assert data == expected

rust/ffi/node/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "lancedb-node"
-version = "0.13.0"
+version = "0.13.1-beta.0"
 description = "Serverless, low-latency vector database for AI applications"
 license.workspace = true
 edition.workspace = true

rust/lancedb/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "lancedb"
-version = "0.13.0"
+version = "0.13.1-beta.0"
 edition.workspace = true
 description = "LanceDB: A serverless, low-latency vector database for AI applications"
 license.workspace = true