Harrison/duckdb (#2064)

Co-authored-by: Trent Hauck <trent@trenthauck.com>
This commit is contained in:
Harrison Chase
2023-03-27 19:51:34 -07:00
committed by GitHub
parent 76ecca4d53
commit f74a1bebf5
4 changed files with 308 additions and 1 deletions

View File

@@ -0,0 +1,56 @@
import unittest
from langchain.document_loaders.duckdb_loader import DuckDBLoader
try:
import duckdb # noqa: F401
duckdb_installed = True
except ImportError:
duckdb_installed = False
@unittest.skipIf(not duckdb_installed, "duckdb not installed")
def test_duckdb_loader_no_options() -> None:
"""Test DuckDB loader."""
loader = DuckDBLoader("SELECT 1 AS a, 2 AS b")
docs = loader.load()
assert len(docs) == 1
assert docs[0].page_content == "a: 1\nb: 2"
assert docs[0].metadata == {}
@unittest.skipIf(not duckdb_installed, "duckdb not installed")
def test_duckdb_loader_page_content_columns() -> None:
"""Test DuckDB loader."""
loader = DuckDBLoader(
"SELECT 1 AS a, 2 AS b UNION SELECT 3 AS a, 4 AS b",
page_content_columns=["a"],
)
docs = loader.load()
assert len(docs) == 2
assert docs[0].page_content == "a: 1"
assert docs[0].metadata == {}
assert docs[1].page_content == "a: 3"
assert docs[1].metadata == {}
@unittest.skipIf(not duckdb_installed, "duckdb not installed")
def test_duckdb_loader_metadata_columns() -> None:
"""Test DuckDB loader."""
loader = DuckDBLoader(
"SELECT 1 AS a, 2 AS b",
page_content_columns=["a"],
metadata_columns=["b"],
)
docs = loader.load()
assert len(docs) == 1
assert docs[0].page_content == "a: 1"
assert docs[0].metadata == {"b": 2}