Implementation:Langchain ai Langchain BaseStoreTests

Knowledge Sources	Langchain_ai_Langchain
Domains	Testing, Key-Value Store, Standard Tests
Last Updated	2026-02-11 00:00 GMT

Overview

Standard integration test suites for validating synchronous and asynchronous implementations of the LangChain BaseStore key-value abstraction.

Description

This module provides two abstract test suite classes -- BaseStoreSyncTests and BaseStoreAsyncTests -- in the langchain-tests (standard-tests) package. They verify that implementations of BaseStore[str, V] from langchain-core correctly support the key-value store API including mget, mset, mdelete, and yield_keys operations. Implementers subclass one or both suites and provide pytest fixtures that return an empty store and three example values.

Usage

Import BaseStoreSyncTests or BaseStoreAsyncTests when writing integration tests for a custom BaseStore implementation to ensure it conforms to the standard key-value store contract.

Code Reference

Source Location

• Repository: Langchain_ai_Langchain
• File: libs/standard-tests/langchain_tests/integration_tests/base_store.py
• Lines: 1-315

Signature

class BaseStoreSyncTests(BaseStandardTests, Generic[V]):
    @abstractmethod
    @pytest.fixture
    def kv_store(self) -> BaseStore[str, V]: ...

    @abstractmethod
    @pytest.fixture
    def three_values(self) -> tuple[V, V, V]: ...

class BaseStoreAsyncTests(BaseStandardTests, Generic[V]):
    @abstractmethod
    @pytest.fixture
    async def kv_store(self) -> BaseStore[str, V]: ...

    @abstractmethod
    @pytest.fixture
    def three_values(self) -> tuple[V, V, V]: ...

Import

from langchain_tests.integration_tests.base_store import (
    BaseStoreSyncTests,
    BaseStoreAsyncTests,
)

I/O Contract

Inputs (Fixtures)

Name	Type	Required	Description
kv_store	BaseStore[str, V]	Yes	An empty key-value store instance to test. Must be empty at the start of each test.
three_values	tuple[V, V, V]	Yes	Three example values of type V that will be used in the tests.

Outputs

Name	Type	Description
Test results	pytest outcomes	Pass/fail results for each standard test method.

Test Methods (BaseStoreSyncTests)

Test Method	Description
test_three_values	Verifies that the fixture provides exactly three values in a tuple.
test_kv_store_is_empty	Verifies the store starts empty (mget returns [None, None, None]).
test_set_and_get_values	Tests mset followed by mget for two key-value pairs.
test_store_still_empty	Verifies the fixture resets the store to empty between tests.
test_delete_values	Tests mdelete removes a single key while preserving others.
test_delete_bulk_values	Tests bulk deletion of multiple keys at once.
test_delete_missing_keys	Verifies deleting non-existent keys does not raise exceptions.
test_set_values_is_idempotent	Verifies setting the same keys twice does not duplicate entries.
test_get_can_get_same_value	Tests retrieving the same key multiple times in a single mget call.
test_overwrite_values_by_key	Tests that mset overwrites existing values for the same key.
test_yield_keys	Tests yield_keys returns all keys and supports prefix filtering.

Test Methods (BaseStoreAsyncTests)

The async suite mirrors the sync suite with equivalent tests using amget, amset, amdelete, and ayield_keys async methods.

Usage Examples

Basic Usage

import pytest
from langchain_core.stores import BaseStore
from langchain_tests.integration_tests.base_store import BaseStoreSyncTests

from my_package.stores import MyKeyValueStore


class TestMyKeyValueStore(BaseStoreSyncTests[str]):
    @pytest.fixture
    def kv_store(self) -> BaseStore[str, str]:
        store = MyKeyValueStore()
        # Ensure the store is empty
        return store

    @pytest.fixture
    def three_values(self) -> tuple[str, str, str]:
        return ("value1", "value2", "value3")