CrateDB: Document Loader

Based on previous contributions, this has effectively become just a
naming-things wrapper around `SQLDatabaseLoader`.

libs/community/extended_testing_deps.txt

@@ -14,6 +14,7 @@ chardet>=5.1.0,<6
 cloudpathlib>=0.18,<0.19
 cloudpickle>=2.0.0
 cohere>=4,<6
+crate==1.0.0dev1
 databricks-vectorsearch>=0.21,<0.22
 datasets>=2.15.0,<3
 dgml-utils>=0.3.0,<0.4
@@ -76,6 +77,7 @@ requests-toolbelt>=1.0.0,<2
 rspace_client>=2.5.0,<3
 scikit-learn>=1.2.2,<2
 simsimd>=5.0.0,<6
+sqlalchemy-cratedb>=0.40.0,<1
 sqlite-vss>=0.1.2,<0.2
 sqlite-vec>=0.1.0,<0.2
 sseclient-py>=1.8.0,<2

libs/community/langchain_community/document_loaders/__init__.py

@@ -129,6 +129,9 @@
     from langchain_community.document_loaders.couchbase import (
         CouchbaseLoader,
     )
+    from langchain_community.document_loaders.cratedb import (
+        CrateDBLoader,
+    )
     from langchain_community.document_loaders.csv_loader import (
         CSVLoader,
         UnstructuredCSVLoader,
@@ -576,6 +579,7 @@
     "ConcurrentLoader": "langchain_community.document_loaders.concurrent",
     "ConfluenceLoader": "langchain_community.document_loaders.confluence",
     "CouchbaseLoader": "langchain_community.document_loaders.couchbase",
+    "CrateDBLoader": "langchain_community.document_loaders.cratedb",
     "CubeSemanticLoader": "langchain_community.document_loaders.cube_semantic",
     "DataFrameLoader": "langchain_community.document_loaders.dataframe",
     "DatadogLogsLoader": "langchain_community.document_loaders.datadog_logs",
@@ -782,6 +786,7 @@ def __getattr__(name: str) -> Any:
     "ConcurrentLoader",
     "ConfluenceLoader",
     "CouchbaseLoader",
+    "CrateDBLoader",
     "CubeSemanticLoader",
     "DataFrameLoader",
     "DatadogLogsLoader",

libs/community/langchain_community/document_loaders/cratedb.py

@@ -0,0 +1,5 @@
+from langchain_community.document_loaders.sql_database import SQLDatabaseLoader
+
+
+class CrateDBLoader(SQLDatabaseLoader):
+    pass

libs/community/tests/examples/mlb_teams_2012.sql

@@ -1,6 +1,7 @@
 -- Provisioning table "mlb_teams_2012".
 --
 -- psql postgresql://postgres@localhost < mlb_teams_2012.sql
+-- crash < mlb_teams_2012.sql
 
 DROP TABLE IF EXISTS mlb_teams_2012;
 CREATE TABLE mlb_teams_2012 ("Team" VARCHAR, "Payroll (millions)" FLOAT, "Wins" BIGINT);

libs/community/tests/integration_tests/document_loaders/docker-compose/cratedb.yml

@@ -0,0 +1,20 @@
+version: "3"
+
+services:
+  postgresql:
+    image: crate/crate:nightly
+    environment:
+      - CRATE_HEAP_SIZE=4g
+    ports:
+      - "4200:4200"
+      - "5432:5432"
+    command: |
+      crate -Cdiscovery.type=single-node
+    healthcheck:
+      test:
+        [
+          "CMD-SHELL",
+          "curl --silent --fail http://localhost:4200/ || exit 1",
+        ]
+      interval: 5s
+      retries: 60

libs/community/tests/integration_tests/document_loaders/docker-compose/postgresql.yml

@@ -0,0 +1,19 @@
+version: "3"
+
+services:
+  postgresql:
+    image: postgres:16
+    environment:
+      - POSTGRES_HOST_AUTH_METHOD=trust
+    ports:
+      - "5432:5432"
+    command: |
+      postgres -c log_statement=all
+    healthcheck:
+      test:
+        [
+          "CMD-SHELL",
+          "psql postgresql://postgres@localhost --command 'SELECT 1;' || exit 1",
+        ]
+      interval: 5s
+      retries: 60

libs/community/tests/integration_tests/document_loaders/test_sql_database.py

@@ -47,6 +47,14 @@
     warnings.warn("psycopg2 not installed, skipping corresponding tests", UserWarning)
     psycopg2_installed = False
 
+try:
+    import sqlalchemy_cratedb  # noqa: F401
+
+    cratedb_installed = True
+except ImportError:
+    warnings.warn("cratedb not installed, skipping corresponding tests", UserWarning)
+    cratedb_installed = False
+
 
 @pytest.fixture()
 def engine(db_uri: str) -> sa.Engine:
@@ -75,6 +83,9 @@ def provision_database(engine: sa.Engine) -> None:
                 continue
             connection.execute(sa.text(statement))
             connection.commit()
+        if engine.dialect.name.startswith("crate"):
+            connection.execute(sa.text("REFRESH TABLE mlb_teams_2012;"))
+            connection.commit()
 
 
 tmpdir = TemporaryDirectory()
@@ -103,6 +114,16 @@ def pytest_generate_tests(metafunc: "Metafunc") -> None:
                 "postgresql+psycopg2://langchain:langchain@localhost:6023/langchain"
             )
             ids.append("postgresql")
+        if cratedb_installed:
+            # We use non-standard port for testing purposes.
+            # The easiest way to spin up the PostgreSQL instance is to use
+            # the docker compose file at the root of the repo located at
+            # langchain/docker/docker-compose.yml
+            # use `docker compose up postgres` to start the instance
+            # it will have the appropriate credentials set up including
+            # being exposed on the appropriate port.
+            urls.append("crate://crate@localhost/?schema=testdrive")
+            ids.append("cratedb")
 
         metafunc.parametrize("db_uri", urls, ids=ids)
 

libs/community/tests/unit_tests/document_loaders/test_imports.py

@@ -48,6 +48,7 @@
     "ConcurrentLoader",
     "ConfluenceLoader",
     "CouchbaseLoader",
+    "CrateDBLoader",
     "CubeSemanticLoader",
     "DataFrameLoader",
     "DatadogLogsLoader",