Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

community: Use astrapy built-in pagination prefetch in AstraDBLoader #17569

Merged
merged 2 commits into from
Feb 15, 2024
Merged

community: Use astrapy built-in pagination prefetch in AstraDBLoader #17569

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
579992e
Use astrapy built-in pagination prefetch in AstraDBLoader
cbornet Feb 15, 2024
7f3399a
Merge branch 'master' into doc-loader-refacto
eyurtsev Feb 15, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
73 changes: 17 additions & 56 deletions libs/community/langchain_community/document_loaders/astradb.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,8 +2,6 @@

import json
import logging
import threading
from queue import Queue
from typing import (
TYPE_CHECKING,
Any,
Expand All @@ -16,7 +14,6 @@
)

from langchain_core.documents import Document
from langchain_core.runnables import run_in_executor

from langchain_community.document_loaders.base import BaseLoader
from langchain_community.utilities.astradb import _AstraDBEnvironment
Expand All @@ -33,6 +30,7 @@ class AstraDBLoader(BaseLoader):
def __init__(
self,
collection_name: str,
*,
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Not related to this PR, but the loader could use a lot of in code documentation. It's surfaced in the APIReference and in interactive environments. Some of the variables are self explanatory, but other variables like filter_criteria or projection would require reading through the code or fishing for documentation somewhere else.

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Sure. I'll add pydoc as follow-up PR.

token: Optional[str] = None,
api_endpoint: Optional[str] = None,
astra_db_client: Optional[AstraDB] = None,
Expand Down Expand Up @@ -65,38 +63,27 @@ def load(self) -> List[Document]:
return list(self.lazy_load())

def lazy_load(self) -> Iterator[Document]:
queue = Queue(self.nb_prefetched) # type: ignore
t = threading.Thread(target=self.fetch_results, args=(queue,))
t.start()
while True:
doc = queue.get()
if doc is None:
break
yield doc
t.join()
for doc in self.collection.paginated_find(
filter=self.filter,
options=self.find_options,
projection=self.projection,
sort=None,
prefetched=self.nb_prefetched,
):
yield Document(
page_content=self.extraction_function(doc),
metadata={
"namespace": self.collection.astra_db.namespace,
"api_endpoint": self.collection.astra_db.base_url,
"collection": self.collection_name,
},
)

async def aload(self) -> List[Document]:
"""Load data into Document objects."""
return [doc async for doc in self.alazy_load()]

async def alazy_load(self) -> AsyncIterator[Document]:
if not self.astra_env.async_astra_db:
iterator = run_in_executor(
None,
self.collection.paginated_find,
filter=self.filter,
options=self.find_options,
projection=self.projection,
sort=None,
prefetched=True,
)
done = object()
while True:
item = await run_in_executor(None, lambda it: next(it, done), iterator)
if item is done:
break
yield item # type: ignore[misc]
return
async_collection = await self.astra_env.async_astra_db.collection(
self.collection_name
)
Expand All @@ -105,7 +92,7 @@ async def alazy_load(self) -> AsyncIterator[Document]:
options=self.find_options,
projection=self.projection,
sort=None,
prefetched=True,
prefetched=self.nb_prefetched,
):
yield Document(
page_content=self.extraction_function(doc),
Expand All @@ -115,29 +102,3 @@ async def alazy_load(self) -> AsyncIterator[Document]:
"collection": self.collection_name,
},
)

def fetch_results(self, queue: Queue): # type: ignore[no-untyped-def]
self.fetch_page_result(queue)
while self.find_options.get("pageState"):
self.fetch_page_result(queue)
queue.put(None)

def fetch_page_result(self, queue: Queue): # type: ignore[no-untyped-def]
res = self.collection.find(
filter=self.filter,
options=self.find_options,
projection=self.projection,
sort=None,
)
self.find_options["pageState"] = res["data"].get("nextPageState")
for doc in res["data"]["documents"]:
queue.put(
Document(
page_content=self.extraction_function(doc),
metadata={
"namespace": self.collection.astra_db.namespace,
"api_endpoint": self.collection.astra_db.base_url,
"collection": self.collection.collection_name,
},
)
)
16 changes: 8 additions & 8 deletions libs/community/tests/integration_tests/document_loaders/test_astradb.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,7 @@
import json
import os
import uuid
from typing import TYPE_CHECKING
from typing import TYPE_CHECKING, AsyncIterator, Iterator

import pytest

Expand All @@ -37,12 +37,12 @@ def _has_env_vars() -> bool:


@pytest.fixture
def astra_db_collection() -> AstraDBCollection:
def astra_db_collection() -> Iterator[AstraDBCollection]:
from astrapy.db import AstraDB

astra_db = AstraDB(
token=ASTRA_DB_APPLICATION_TOKEN,
api_endpoint=ASTRA_DB_API_ENDPOINT,
token=ASTRA_DB_APPLICATION_TOKEN or "",
api_endpoint=ASTRA_DB_API_ENDPOINT or "",
namespace=ASTRA_DB_KEYSPACE,
)
collection_name = f"lc_test_loader_{str(uuid.uuid4()).split('-')[0]}"
Expand All @@ -58,12 +58,12 @@ def astra_db_collection() -> AstraDBCollection:


@pytest.fixture
async def async_astra_db_collection() -> AsyncAstraDBCollection:
async def async_astra_db_collection() -> AsyncIterator[AsyncAstraDBCollection]:
from astrapy.db import AsyncAstraDB

astra_db = AsyncAstraDB(
token=ASTRA_DB_APPLICATION_TOKEN,
api_endpoint=ASTRA_DB_API_ENDPOINT,
token=ASTRA_DB_APPLICATION_TOKEN or "",
api_endpoint=ASTRA_DB_API_ENDPOINT or "",
namespace=ASTRA_DB_KEYSPACE,
)
collection_name = f"lc_test_loader_{str(uuid.uuid4()).split('-')[0]}"
Expand Down Expand Up @@ -167,5 +167,5 @@ async def test_extraction_function_async(
find_options={"limit": 30},
extraction_function=lambda x: x["foo"],
)
doc = await anext(loader.alazy_load()) # type: ignore[name-defined]
doc = await loader.alazy_load().__anext__()
assert doc.page_content == "bar"
Loading