You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
langchain/libs/community/tests/integration_tests/cache/test_astradb.py

160 lines
5.7 KiB
Python

"""
Test AstraDB caches. Requires an Astra DB vector instance.
Required to run this test:
- a recent `astrapy` Python package available
- an Astra DB instance;
- the two environment variables set:
export ASTRA_DB_API_ENDPOINT="https://<DB-ID>-us-east1.apps.astra.datastax.com"
export ASTRA_DB_APPLICATION_TOKEN="AstraCS:........."
- optionally this as well (otherwise defaults are used):
export ASTRA_DB_KEYSPACE="my_keyspace"
"""
import os
from typing import AsyncIterator, Iterator
import pytest
from langchain.globals import get_llm_cache, set_llm_cache
from langchain_core.caches import BaseCache
from langchain_core.language_models import LLM
from langchain_core.outputs import Generation, LLMResult
from langchain_community.cache import AstraDBCache, AstraDBSemanticCache
from langchain_community.utilities.astradb import SetupMode
from tests.integration_tests.cache.fake_embeddings import FakeEmbeddings
from tests.unit_tests.llms.fake_llm import FakeLLM
def _has_env_vars() -> bool:
return all(
[
"ASTRA_DB_APPLICATION_TOKEN" in os.environ,
"ASTRA_DB_API_ENDPOINT" in os.environ,
]
)
@pytest.fixture(scope="module")
def astradb_cache() -> Iterator[AstraDBCache]:
cache = AstraDBCache(
collection_name="lc_integration_test_cache",
token=os.environ["ASTRA_DB_APPLICATION_TOKEN"],
api_endpoint=os.environ["ASTRA_DB_API_ENDPOINT"],
namespace=os.environ.get("ASTRA_DB_KEYSPACE"),
)
yield cache
cache.collection.astra_db.delete_collection("lc_integration_test_cache")
@pytest.fixture
async def async_astradb_cache() -> AsyncIterator[AstraDBCache]:
cache = AstraDBCache(
collection_name="lc_integration_test_cache_async",
token=os.environ["ASTRA_DB_APPLICATION_TOKEN"],
api_endpoint=os.environ["ASTRA_DB_API_ENDPOINT"],
namespace=os.environ.get("ASTRA_DB_KEYSPACE"),
setup_mode=SetupMode.ASYNC,
)
yield cache
await cache.async_collection.astra_db.delete_collection(
"lc_integration_test_cache_async"
)
@pytest.fixture(scope="module")
def astradb_semantic_cache() -> Iterator[AstraDBSemanticCache]:
fake_embe = FakeEmbeddings()
sem_cache = AstraDBSemanticCache(
collection_name="lc_integration_test_sem_cache",
token=os.environ["ASTRA_DB_APPLICATION_TOKEN"],
api_endpoint=os.environ["ASTRA_DB_API_ENDPOINT"],
namespace=os.environ.get("ASTRA_DB_KEYSPACE"),
embedding=fake_embe,
)
yield sem_cache
sem_cache.collection.astra_db.delete_collection("lc_integration_test_sem_cache")
@pytest.fixture
async def async_astradb_semantic_cache() -> AsyncIterator[AstraDBSemanticCache]:
fake_embe = FakeEmbeddings()
sem_cache = AstraDBSemanticCache(
collection_name="lc_integration_test_sem_cache_async",
token=os.environ["ASTRA_DB_APPLICATION_TOKEN"],
api_endpoint=os.environ["ASTRA_DB_API_ENDPOINT"],
namespace=os.environ.get("ASTRA_DB_KEYSPACE"),
embedding=fake_embe,
setup_mode=SetupMode.ASYNC,
)
yield sem_cache
sem_cache.collection.astra_db.delete_collection(
"lc_integration_test_sem_cache_async"
)
@pytest.mark.requires("astrapy")
@pytest.mark.skipif(not _has_env_vars(), reason="Missing Astra DB env. vars")
class TestAstraDBCaches:
def test_astradb_cache(self, astradb_cache: AstraDBCache) -> None:
self.do_cache_test(FakeLLM(), astradb_cache, "foo")
async def test_astradb_cache_async(self, async_astradb_cache: AstraDBCache) -> None:
await self.ado_cache_test(FakeLLM(), async_astradb_cache, "foo")
def test_astradb_semantic_cache(
self, astradb_semantic_cache: AstraDBSemanticCache
) -> None:
llm = FakeLLM()
self.do_cache_test(llm, astradb_semantic_cache, "bar")
output = llm.generate(["bar"]) # 'fizz' is erased away now
assert output != LLMResult(
generations=[[Generation(text="fizz")]],
llm_output={},
)
astradb_semantic_cache.clear()
async def test_astradb_semantic_cache_async(
self, async_astradb_semantic_cache: AstraDBSemanticCache
) -> None:
llm = FakeLLM()
await self.ado_cache_test(llm, async_astradb_semantic_cache, "bar")
output = await llm.agenerate(["bar"]) # 'fizz' is erased away now
assert output != LLMResult(
generations=[[Generation(text="fizz")]],
llm_output={},
)
await async_astradb_semantic_cache.aclear()
@staticmethod
def do_cache_test(llm: LLM, cache: BaseCache, prompt: str) -> None:
set_llm_cache(cache)
params = llm.dict()
params["stop"] = None
llm_string = str(sorted([(k, v) for k, v in params.items()]))
get_llm_cache().update("foo", llm_string, [Generation(text="fizz")])
output = llm.generate([prompt])
expected_output = LLMResult(
generations=[[Generation(text="fizz")]],
llm_output={},
)
assert output == expected_output
# clear the cache
cache.clear()
@staticmethod
async def ado_cache_test(llm: LLM, cache: BaseCache, prompt: str) -> None:
set_llm_cache(cache)
params = llm.dict()
params["stop"] = None
llm_string = str(sorted([(k, v) for k, v in params.items()]))
await get_llm_cache().aupdate("foo", llm_string, [Generation(text="fizz")])
output = await llm.agenerate([prompt])
expected_output = LLMResult(
generations=[[Generation(text="fizz")]],
llm_output={},
)
assert output == expected_output
# clear the cache
await cache.aclear()