Fix ord-schema tests (#716)

* skip tests if rdkit cartridge is not available

* message

* skip

* fix notebook

* postgresql version

* apt

* apt-get

* notebook

* conda-forge

.github/workflows/run_tests.yml

@@ -33,7 +33,8 @@ jobs:
         shell: bash -l {0}
         run: |
           # NOTE(skearnes): conda is only used for postgres (not python).
-          conda install -c rdkit rdkit-postgresql==2020.03.3.0
+          # NOTE(skearnes): rdkit-postgresql may not be available for ARM.
+          conda install -c rdkit rdkit-postgresql==2020.03.3.0 || conda install -c conda-forge postgresql
           initdb
       - uses: actions/setup-python@v4
         with:

ord_schema/orm/database.py

@@ -20,7 +20,7 @@
 from sqlalchemy import cast, delete, func, select, text, update
 from sqlalchemy.dialects.postgresql import insert
 from sqlalchemy.engine import Engine
-from sqlalchemy.exc import OperationalError
+from sqlalchemy.exc import OperationalError, NotSupportedError
 from sqlalchemy.orm import Session
 
 from ord_schema.logging import get_logger
@@ -55,14 +55,16 @@ def prepare_database(engine: Engine) -> bool:
     with engine.begin() as connection:
         connection.execute(text("CREATE SCHEMA IF NOT EXISTS ord"))
         connection.execute(text("CREATE SCHEMA IF NOT EXISTS rdkit"))
-        try:
+    try:
+        with engine.begin() as connection:
             # NOTE(skearnes): The RDKit PostgreSQL extension works best in the public schema.
             connection.execute(text("CREATE EXTENSION IF NOT EXISTS rdkit"))
-            rdkit_cartridge = True
-        except OperationalError:
+        rdkit_cartridge = True
+    except (OperationalError, NotSupportedError):
+        with engine.begin() as connection:
             logger.warning("RDKit PostgreSQL cartridge is not installed; structure search will be disabled")
             connection.execute(text("CREATE EXTENSION IF NOT EXISTS btree_gist"))
-            rdkit_cartridge = False
+        rdkit_cartridge = False
     with patch.dict(os.environ, {"ORD_POSTGRES_RDKIT": "1" if rdkit_cartridge else "0"}):
         Base.metadata.create_all(engine)
     return rdkit_cartridge

ord_schema/orm/rdkit_mappers_test.py

@@ -15,30 +15,37 @@
 """Tests for ord_schema.orm.rdkit_mappers."""
 import pytest
 from sqlalchemy import select
+from sqlalchemy.exc import ProgrammingError
 from ord_schema.orm.mappers import Mappers
 from ord_schema.orm.rdkit_mappers import FingerprintType, RDKitMol
 
 
 def test_tanimoto_operator(test_session):
-    query = (
-        select(Mappers.Reaction)
-        .join(Mappers.ReactionInput)
-        .join(Mappers.Compound)
-        .join(RDKitMol)
-        .where(RDKitMol.morgan_bfp % FingerprintType.MORGAN_BFP("c1ccccc1CCC(O)C"))
-    )
-    results = test_session.execute(query)
-    assert len(results.fetchall()) == 20
+    try:
+        query = (
+            select(Mappers.Reaction)
+            .join(Mappers.ReactionInput)
+            .join(Mappers.Compound)
+            .join(RDKitMol)
+            .where(RDKitMol.morgan_bfp % FingerprintType.MORGAN_BFP("c1ccccc1CCC(O)C"))
+        )
+        results = test_session.execute(query)
+        assert len(results.fetchall()) == 20
+    except ProgrammingError as error:
+        pytest.skip(f"RDKit cartridge is required: {error}")
 
 
 @pytest.mark.parametrize("fp_type", list(FingerprintType))
 def test_tanimoto(test_session, fp_type):
-    query = (
-        select(Mappers.Reaction)
-        .join(Mappers.ReactionInput)
-        .join(Mappers.Compound)
-        .join(RDKitMol)
-        .where(RDKitMol.tanimoto("c1ccccc1CCC(O)C", fp_type=fp_type) > 0.5)
-    )
-    results = test_session.execute(query)
-    assert len(results.fetchall()) == 20
+    try:
+        query = (
+            select(Mappers.Reaction)
+            .join(Mappers.ReactionInput)
+            .join(Mappers.Compound)
+            .join(RDKitMol)
+            .where(RDKitMol.tanimoto("c1ccccc1CCC(O)C", fp_type=fp_type) > 0.5)
+        )
+        results = test_session.execute(query)
+        assert len(results.fetchall()) == 20
+    except ProgrammingError as error:
+        pytest.skip(f"RDKit cartridge is required: {error}")

ord_schema/orm/scripts/add_datasets_test.py

@@ -16,6 +16,7 @@
 import os
 
 import docopt
+import pytest
 from sqlalchemy import create_engine
 from testing.postgresql import Postgresql
 
@@ -26,7 +27,8 @@
 def test_main():
     with Postgresql() as postgres:
         engine = create_engine(postgres.url(), future=True)
-        assert prepare_database(engine)  # Requires RDKit.
+        if not prepare_database(engine):
+            pytest.skip("RDKit cartridge is required")
         argv = [
             "--url",
             postgres.url(),