Files
lancedb/docs/web-tests/py/test_connection.py
T
Jack Ye 564474f51a feat: use SQL for remote function lifecycle
Remote function registration, exact-version lookup, listing, and drop now execute SQL through the Flight client. Registration waits for the durable image build and catalog publication before yielding its typed `FunctionVersion`. SQL also covers source-inclusive description and pinned metadata updates.

Registration and drop handles use `job_type="query"`; the single `job.id` is the SQL query UUID and can be inspected with `describe_query()` on the submitting connection. Waiting drains the SQL result once and caches completion for repeated waits. The drop boolean reports catalog removal; physical cleanup continues separately. Query handles have no registry history and cannot be reopened with `open_job()`.
2026-10-30 22:35:35 +08:00

229 lines
7.1 KiB
Python

# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright The LanceDB Authors
from pathlib import Path
import shutil
import pytest
def test_connection():
# --8<-- [start:connect]
import lancedb
uri = "ex_lancedb"
db = lancedb.connect(uri)
# --8<-- [end:connect]
assert db is not None
shutil.rmtree(uri, ignore_errors=True)
assert not Path(uri).exists()
async def connect_async_example():
# --8<-- [start:connect_async]
import lancedb
uri = "ex_lancedb"
async_db = await lancedb.connect_async(uri)
# --8<-- [end:connect_async]
return async_db
def connect_enterprise_quickstart_config():
# --8<-- [start:connect_enterprise_quickstart]
import os
import lancedb
db = lancedb.connect(
uri=os.environ["LANCEDB_URI"], # db://your-database
api_key=os.environ["LANCEDB_API_KEY"],
region=os.environ.get("LANCEDB_REGION", "us-east-1"),
host_override=os.environ["LANCEDB_HOST_OVERRIDE"], # https://your-endpoint
sql_host_override=os.environ["LANCEDB_SQL_HOST_OVERRIDE"],
)
# --8<-- [end:connect_enterprise_quickstart]
return db
def test_connect_enterprise_quickstart(monkeypatch):
import lancedb
captured = {}
def fake_connect(**kwargs):
captured.update(kwargs)
return object()
monkeypatch.setattr(lancedb, "connect", fake_connect)
monkeypatch.setenv("LANCEDB_URI", "db://your-database")
monkeypatch.setenv("LANCEDB_API_KEY", "your-api-key")
monkeypatch.delenv("LANCEDB_REGION", raising=False)
monkeypatch.setenv("LANCEDB_HOST_OVERRIDE", "https://your-endpoint")
monkeypatch.setenv(
"LANCEDB_SQL_HOST_OVERRIDE", "grpc+tls://your-sql-endpoint:10026"
)
assert connect_enterprise_quickstart_config() is not None
assert captured == {
"uri": "db://your-database",
"api_key": "your-api-key",
"region": "us-east-1",
"host_override": "https://your-endpoint",
"sql_host_override": "grpc+tls://your-sql-endpoint:10026",
}
def test_connect_enterprise_quickstart_requires_an_endpoint(monkeypatch):
monkeypatch.setenv("LANCEDB_URI", "db://your-database")
monkeypatch.setenv("LANCEDB_API_KEY", "your-api-key")
monkeypatch.delenv("LANCEDB_HOST_OVERRIDE", raising=False)
with pytest.raises(KeyError, match="LANCEDB_HOST_OVERRIDE"):
connect_enterprise_quickstart_config()
def test_connect_enterprise_quickstart_requires_a_sql_endpoint(monkeypatch):
monkeypatch.setenv("LANCEDB_URI", "db://your-database")
monkeypatch.setenv("LANCEDB_API_KEY", "your-api-key")
monkeypatch.setenv("LANCEDB_HOST_OVERRIDE", "https://your-endpoint")
monkeypatch.delenv("LANCEDB_SQL_HOST_OVERRIDE", raising=False)
with pytest.raises(KeyError, match="LANCEDB_SQL_HOST_OVERRIDE"):
connect_enterprise_quickstart_config()
def connect_object_storage_config():
# --8<-- [start:connect_object_storage]
import lancedb
uri = "s3://your-bucket/path"
# You can also use "gs://your-bucket/path" or "az://your-container/path".
db = lancedb.connect(uri)
# --8<-- [end:connect_object_storage]
return db
def namespace_table_ops_example():
# --8<-- [start:namespace_table_ops]
import lancedb
db = lancedb.connect_namespace("dir", {"root": "./local_lancedb"})
# Create namespace tree: prod/search
db.create_namespace(["prod"], mode="exist_ok")
db.create_namespace(["prod", "search"], mode="exist_ok")
db.create_namespace(["prod", "recommendations"], mode="exist_ok")
db.create_table(
"user",
data=[{"id": 1, "vector": [0.1, 0.2], "name": "alice"}],
namespace_path=["prod", "search"],
mode="create", # use "overwrite" only if you want to replace existing table
)
db.create_table(
"user",
data=[{"id": 2, "vector": [0.3, 0.4], "name": "bob"}],
namespace_path=["prod", "recommendations"],
mode="create", # use "overwrite" only if you want to replace existing table
)
# Verify: listings return a response with the names and a page token
print(db.list_namespaces().namespaces) # ['prod']
print(db.list_namespaces(namespace_path=["prod"]).namespaces)
# ['recommendations', 'search']
print(db.list_tables(namespace_path=["prod", "search"]).tables) # ['user']
print(db.list_tables(namespace_path=["prod", "recommendations"]).tables) # ['user']
# --8<-- [end:namespace_table_ops]
def test_namespace_table_ops(tmp_path, monkeypatch, capsys):
monkeypatch.chdir(tmp_path)
namespace_table_ops_example()
assert capsys.readouterr().out.splitlines() == [
"['prod']",
"['recommendations', 'search']",
"['user']",
"['user']",
]
def namespace_admin_ops_example():
# --8<-- [start:namespace_admin_ops]
import lancedb
db = lancedb.connect_namespace("dir", {"root": "./local_lancedb"})
namespace = ["prod", "search"]
db.create_namespace(["prod"])
db.create_namespace(["prod", "search"])
child_namespaces = db.list_namespaces(namespace_path=["prod"]).namespaces
print(f"Child namespaces under ['prod']: {child_namespaces}")
# Child namespaces under ['prod']: ['search']
metadata = db.describe_namespace(["prod", "search"])
print(f"Metadata for namespace {namespace}: {metadata}")
# Metadata for namespace ['prod', 'search']: properties=None
db.drop_namespace(["prod", "search"], mode="skip")
db.drop_namespace(["prod"], mode="skip")
# --8<-- [end:namespace_admin_ops]
return db, child_namespaces, metadata
def test_namespace_admin_ops(tmp_path, monkeypatch, capsys):
monkeypatch.chdir(tmp_path)
db, child_namespaces, metadata = namespace_admin_ops_example()
assert child_namespaces == ["search"]
assert capsys.readouterr().out.splitlines()[0] == (
"Child namespaces under ['prod']: ['search']"
)
assert db.list_namespaces().namespaces == []
async def connect_object_storage_config_async():
# --8<-- [start:connect_object_storage_async]
import lancedb
uri = "s3://your-bucket/path"
# You can also use "gs://your-bucket/path" or "az://your-container/path".
async_db = await lancedb.connect_async(uri)
# --8<-- [end:connect_object_storage_async]
return async_db
def test_connect_enterprise_placeholders(monkeypatch):
import lancedb
captured = {}
def fake_connect(**kwargs):
captured.update(kwargs)
return object()
monkeypatch.setattr(lancedb, "connect", fake_connect)
# --8<-- [start:connect_enterprise_placeholders]
import lancedb
db = lancedb.connect(
uri="db://your-database",
api_key="your-api-key",
region="us-east-1",
host_override="https://your-endpoint",
sql_host_override="grpc+tls://your-sql-endpoint:10026",
)
# --8<-- [end:connect_enterprise_placeholders]
assert db is not None
assert captured == {
"uri": "db://your-database",
"api_key": "your-api-key",
"region": "us-east-1",
"host_override": "https://your-endpoint",
"sql_host_override": "grpc+tls://your-sql-endpoint:10026",
}