cognee 0.4.1__py3-none-any.whl → 0.5.0__py3-none-any.whl

cognee/tests/unit/api/test_ontology_endpoint.py

@@ -0,0 +1,252 @@
+import pytest
+import uuid
+from fastapi.testclient import TestClient
+from unittest.mock import Mock
+from types import SimpleNamespace
+from cognee.api.client import app
+from cognee.modules.users.methods import get_authenticated_user
+
+
+@pytest.fixture(scope="session")
+def test_client():
+    # Keep a single TestClient (and event loop) for the whole module.
+    # Re-creating TestClient repeatedly can break async DB connections (asyncpg loop mismatch).
+    with TestClient(app) as c:
+        yield c
+
+
+@pytest.fixture
+def client(test_client, mock_default_user):
+    async def override_get_authenticated_user():
+        return mock_default_user
+
+    app.dependency_overrides[get_authenticated_user] = override_get_authenticated_user
+    yield test_client
+    app.dependency_overrides.pop(get_authenticated_user, None)
+
+
+@pytest.fixture
+def mock_user():
+    user = Mock()
+    user.id = "test-user-123"
+    return user
+
+
+@pytest.fixture
+def mock_default_user():
+    """Mock default user for testing."""
+    return SimpleNamespace(
+        id=str(uuid.uuid4()),
+        email="default@example.com",
+        is_active=True,
+        tenant_id=str(uuid.uuid4()),
+    )
+
+
+def test_upload_ontology_success(client):
+    """Test successful ontology upload"""
+    ontology_content = (
+        b"<rdf:RDF xmlns:rdf='http://www.w3.org/1999/02/22-rdf-syntax-ns#'></rdf:RDF>"
+    )
+    unique_key = f"test_ontology_{uuid.uuid4().hex[:8]}"
+
+    response = client.post(
+        "/api/v1/ontologies",
+        files=[("ontology_file", ("test.owl", ontology_content, "application/xml"))],
+        data={"ontology_key": unique_key, "description": "Test"},
+    )
+
+    assert response.status_code == 200
+    data = response.json()
+    assert data["uploaded_ontologies"][0]["ontology_key"] == unique_key
+    assert "uploaded_at" in data["uploaded_ontologies"][0]
+
+
+def test_upload_ontology_invalid_file(client):
+    """Test 400 response for non-.owl files"""
+    unique_key = f"test_ontology_{uuid.uuid4().hex[:8]}"
+    response = client.post(
+        "/api/v1/ontologies",
+        files={"ontology_file": ("test.txt", b"not xml")},
+        data={"ontology_key": unique_key},
+    )
+    assert response.status_code == 400
+
+
+def test_upload_ontology_missing_data(client):
+    """Test 400 response for missing file or key"""
+    # Missing file
+    response = client.post("/api/v1/ontologies", data={"ontology_key": "test"})
+    assert response.status_code == 400
+
+    # Missing key
+    response = client.post(
+        "/api/v1/ontologies", files=[("ontology_file", ("test.owl", b"xml", "application/xml"))]
+    )
+    assert response.status_code == 400
+
+
+def test_upload_ontology_without_auth_header(client):
+    """Test behavior when no explicit authentication header is provided."""
+    unique_key = f"test_ontology_{uuid.uuid4().hex[:8]}"
+    response = client.post(
+        "/api/v1/ontologies",
+        files=[("ontology_file", ("test.owl", b"<rdf></rdf>", "application/xml"))],
+        data={"ontology_key": unique_key},
+    )
+
+    assert response.status_code == 200
+    data = response.json()
+    assert data["uploaded_ontologies"][0]["ontology_key"] == unique_key
+    assert "uploaded_at" in data["uploaded_ontologies"][0]
+
+
+def test_upload_multiple_ontologies_in_single_request_is_rejected(client):
+    """Uploading multiple ontology files in a single request should fail."""
+    import io
+
+    file1_content = b"<rdf:RDF xmlns:rdf='http://www.w3.org/1999/02/22-rdf-syntax-ns#'></rdf:RDF>"
+    file2_content = b"<rdf:RDF xmlns:rdf='http://www.w3.org/1999/02/22-rdf-syntax-ns#'></rdf:RDF>"
+
+    files = [
+        ("ontology_file", ("vehicles.owl", io.BytesIO(file1_content), "application/xml")),
+        ("ontology_file", ("manufacturers.owl", io.BytesIO(file2_content), "application/xml")),
+    ]
+    data = {"ontology_key": "vehicles", "description": "Base vehicles"}
+
+    response = client.post("/api/v1/ontologies", files=files, data=data)
+
+    assert response.status_code == 400
+    assert "Only one ontology_file is allowed" in response.json()["error"]
+
+
+def test_upload_endpoint_rejects_array_style_fields(client):
+    """Array-style form values should be rejected (no backwards compatibility)."""
+    import io
+    import json
+
+    file_content = b"<rdf:RDF xmlns:rdf='http://www.w3.org/1999/02/22-rdf-syntax-ns#'></rdf:RDF>"
+
+    files = [("ontology_file", ("single.owl", io.BytesIO(file_content), "application/xml"))]
+    data = {
+        "ontology_key": json.dumps(["single_key"]),
+        "description": json.dumps(["Single ontology"]),
+    }
+
+    response = client.post("/api/v1/ontologies", files=files, data=data)
+
+    assert response.status_code == 400
+    assert "ontology_key must be a string" in response.json()["error"]
+
+
+def test_cognify_with_multiple_ontologies(client):
+    """Test cognify endpoint accepts multiple ontology keys"""
+    payload = {
+        "datasets": ["test_dataset"],
+        "ontology_key": ["ontology1", "ontology2"],  # Array instead of string
+        "run_in_background": False,
+    }
+
+    response = client.post("/api/v1/cognify", json=payload)
+
+    # Should not fail due to ontology_key type
+    assert response.status_code in [200, 400, 409]  # May fail for other reasons, not type
+
+
+def test_complete_multifile_workflow(client):
+    """Test workflow: upload ontologies one-by-one → cognify with multiple keys"""
+    import io
+
+    # Step 1: Upload two ontologies (one-by-one)
+    file1_content = b"""<?xml version="1.0"?>
+    <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+             xmlns:owl="http://www.w3.org/2002/07/owl#">
+        <owl:Class rdf:ID="Vehicle"/>
+    </rdf:RDF>"""
+
+    file2_content = b"""<?xml version="1.0"?>
+    <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+             xmlns:owl="http://www.w3.org/2002/07/owl#">
+        <owl:Class rdf:ID="Manufacturer"/>
+    </rdf:RDF>"""
+
+    upload_response_1 = client.post(
+        "/api/v1/ontologies",
+        files=[("ontology_file", ("vehicles.owl", io.BytesIO(file1_content), "application/xml"))],
+        data={"ontology_key": "vehicles", "description": "Vehicle ontology"},
+    )
+    assert upload_response_1.status_code == 200
+
+    upload_response_2 = client.post(
+        "/api/v1/ontologies",
+        files=[
+            ("ontology_file", ("manufacturers.owl", io.BytesIO(file2_content), "application/xml"))
+        ],
+        data={"ontology_key": "manufacturers", "description": "Manufacturer ontology"},
+    )
+    assert upload_response_2.status_code == 200
+
+    # Step 2: Verify ontologies are listed
+    list_response = client.get("/api/v1/ontologies")
+    assert list_response.status_code == 200
+    ontologies = list_response.json()
+    assert "vehicles" in ontologies
+    assert "manufacturers" in ontologies
+
+    # Step 3: Test cognify with multiple ontologies
+    cognify_payload = {
+        "datasets": ["test_dataset"],
+        "ontology_key": ["vehicles", "manufacturers"],
+        "run_in_background": False,
+    }
+
+    cognify_response = client.post("/api/v1/cognify", json=cognify_payload)
+    # Should not fail due to ontology handling (may fail for dataset reasons)
+    assert cognify_response.status_code != 400  # Not a validation error
+
+
+def test_upload_error_handling(client):
+    """Test error handling for invalid uploads (single-file endpoint)."""
+    import io
+    import json
+
+    # Array-style key should be rejected
+    file_content = b"<rdf:RDF></rdf:RDF>"
+    files = [("ontology_file", ("test.owl", io.BytesIO(file_content), "application/xml"))]
+    data = {
+        "ontology_key": json.dumps(["key1", "key2"]),
+        "description": "desc1",
+    }
+
+    response = client.post("/api/v1/ontologies", files=files, data=data)
+    assert response.status_code == 400
+    assert "ontology_key must be a string" in response.json()["error"]
+
+    # Duplicate key should be rejected
+    response_1 = client.post(
+        "/api/v1/ontologies",
+        files=[("ontology_file", ("test1.owl", io.BytesIO(file_content), "application/xml"))],
+        data={"ontology_key": "duplicate", "description": "desc1"},
+    )
+    assert response_1.status_code == 200
+
+    response_2 = client.post(
+        "/api/v1/ontologies",
+        files=[("ontology_file", ("test2.owl", io.BytesIO(file_content), "application/xml"))],
+        data={"ontology_key": "duplicate", "description": "desc2"},
+    )
+    assert response_2.status_code == 400
+    assert "already exists" in response_2.json()["error"]
+
+
+def test_cognify_missing_ontology_key(client):
+    """Test cognify with non-existent ontology key"""
+    payload = {
+        "datasets": ["test_dataset"],
+        "ontology_key": ["nonexistent_key"],
+        "run_in_background": False,
+    }
+
+    response = client.post("/api/v1/cognify", json=payload)
+    assert response.status_code == 409
+    assert "Ontology key 'nonexistent_key' not found" in response.json()["error"]

cognee/tests/unit/infrastructure/databases/cache/test_cache_config.py

@@ -8,6 +8,7 @@ def test_cache_config_defaults():
     """Test that CacheConfig has the correct default values."""
     config = CacheConfig()
 
+    assert config.cache_backend == "fs"
     assert config.caching is False
     assert config.shared_kuzu_lock is False
     assert config.cache_host == "localhost"
@@ -19,6 +20,7 @@ def test_cache_config_defaults():
 def test_cache_config_custom_values():
     """Test that CacheConfig accepts custom values."""
     config = CacheConfig(
+        cache_backend="redis",
         caching=True,
         shared_kuzu_lock=True,
         cache_host="redis.example.com",
@@ -27,6 +29,7 @@ def test_cache_config_custom_values():
         agentic_lock_timeout=180,
     )
 
+    assert config.cache_backend == "redis"
     assert config.caching is True
     assert config.shared_kuzu_lock is True
     assert config.cache_host == "redis.example.com"
@@ -38,6 +41,7 @@ def test_cache_config_custom_values():
 def test_cache_config_to_dict():
     """Test the to_dict method returns all configuration values."""
     config = CacheConfig(
+        cache_backend="fs",
         caching=True,
         shared_kuzu_lock=True,
         cache_host="test-host",
@@ -49,6 +53,7 @@ def test_cache_config_to_dict():
     config_dict = config.to_dict()
 
     assert config_dict == {
+        "cache_backend": "fs",
         "caching": True,
         "shared_kuzu_lock": True,
         "cache_host": "test-host",

cognee/tests/unit/infrastructure/databases/test_index_data_points.py

@@ -0,0 +1,27 @@
+import pytest
+from unittest.mock import AsyncMock, patch, MagicMock
+from cognee.tasks.storage.index_data_points import index_data_points
+from cognee.infrastructure.engine import DataPoint
+
+
+class TestDataPoint(DataPoint):
+    name: str
+    metadata: dict = {"index_fields": ["name"]}
+
+
+@pytest.mark.asyncio
+async def test_index_data_points_calls_vector_engine():
+    """Test that index_data_points creates vector index and indexes data."""
+    data_points = [TestDataPoint(name="test1")]
+
+    mock_vector_engine = AsyncMock()
+    mock_vector_engine.embedding_engine.get_batch_size = MagicMock(return_value=100)
+
+    with patch.dict(
+        index_data_points.__globals__,
+        {"get_vector_engine": lambda: mock_vector_engine},
+    ):
+        await index_data_points(data_points)
+
+    assert mock_vector_engine.create_vector_index.await_count >= 1
+    assert mock_vector_engine.index_data_points.await_count >= 1

cognee/tests/unit/infrastructure/databases/test_index_graph_edges.py

@@ -5,8 +5,7 @@ from cognee.tasks.storage.index_graph_edges import index_graph_edges
 
 @pytest.mark.asyncio
 async def test_index_graph_edges_success():
-    """Test that index_graph_edges uses the index datapoints and creates vector index."""
-    # Create the mocks for the graph and vector engines.
+    """Test that index_graph_edges retrieves edges and delegates to index_data_points."""
     mock_graph_engine = AsyncMock()
     mock_graph_engine.get_graph_data.return_value = (
         None,
@@ -15,26 +14,23 @@ async def test_index_graph_edges_success():
             [{"relationship_name": "rel2"}],
         ],
     )
-    mock_vector_engine = AsyncMock()
-    mock_vector_engine.embedding_engine.get_batch_size = MagicMock(return_value=100)
+    mock_index_data_points = AsyncMock()
 
-    # Patch the globals of the function so that when it does:
-    #   vector_engine = get_vector_engine()
-    #   graph_engine = await get_graph_engine()
-    # it uses the mocked versions.
     with patch.dict(
         index_graph_edges.__globals__,
         {
             "get_graph_engine": AsyncMock(return_value=mock_graph_engine),
-            "get_vector_engine": lambda: mock_vector_engine,
+            "index_data_points": mock_index_data_points,
         },
     ):
         await index_graph_edges()
 
-    # Assertions on the mock calls.
     mock_graph_engine.get_graph_data.assert_awaited_once()
-    assert mock_vector_engine.create_vector_index.await_count == 1
-    assert mock_vector_engine.index_data_points.await_count == 1
+    mock_index_data_points.assert_awaited_once()
+
+    call_args = mock_index_data_points.call_args[0][0]
+    assert len(call_args) == 2
+    assert all(hasattr(item, "relationship_name") for item in call_args)
 
 
 @pytest.mark.asyncio
@@ -42,20 +38,22 @@ async def test_index_graph_edges_no_relationships():
     """Test that index_graph_edges handles empty relationships correctly."""
     mock_graph_engine = AsyncMock()
     mock_graph_engine.get_graph_data.return_value = (None, [])
-    mock_vector_engine = AsyncMock()
+    mock_index_data_points = AsyncMock()
 
     with patch.dict(
         index_graph_edges.__globals__,
         {
             "get_graph_engine": AsyncMock(return_value=mock_graph_engine),
-            "get_vector_engine": lambda: mock_vector_engine,
+            "index_data_points": mock_index_data_points,
         },
     ):
         await index_graph_edges()
 
     mock_graph_engine.get_graph_data.assert_awaited_once()
-    mock_vector_engine.create_vector_index.assert_not_awaited()
-    mock_vector_engine.index_data_points.assert_not_awaited()
+    mock_index_data_points.assert_awaited_once()
+
+    call_args = mock_index_data_points.call_args[0][0]
+    assert len(call_args) == 0
 
 
 @pytest.mark.asyncio

cognee/tests/unit/infrastructure/llm/test_llm_config.py

@@ -0,0 +1,46 @@
+import pytest
+
+from cognee.infrastructure.llm.config import LLMConfig
+
+
+def test_strip_quotes_from_strings():
+    """
+    Test if the LLMConfig.strip_quotes_from_strings model validator behaves as expected.
+    """
+    config = LLMConfig(
+        # Strings with surrounding double quotes ("value" → value)
+        llm_api_key='"double_value"',
+        # Strings with surrounding single quotes ('value' → value)
+        llm_endpoint="'single_value'",
+        # Strings without quotes (value → value)
+        llm_api_version="no_quotes_value",
+        # Empty quoted strings ("" → empty string)
+        fallback_model='""',
+        # None values (should remain None)
+        baml_llm_api_key=None,
+        # Mixed quotes ("value' → unchanged)
+        fallback_endpoint="\"mixed_quote'",
+        # Strings with internal quotes ("internal\"quotes" → internal"quotes")
+        baml_llm_model='"internal"quotes"',
+    )
+
+    # Strings with surrounding double quotes ("value" → value)
+    assert config.llm_api_key == "double_value"
+
+    # Strings with surrounding single quotes ('value' → value)
+    assert config.llm_endpoint == "single_value"
+
+    # Strings without quotes (value → value)
+    assert config.llm_api_version == "no_quotes_value"
+
+    # Empty quoted strings ("" → empty string)
+    assert config.fallback_model == ""
+
+    # None values (should remain None)
+    assert config.baml_llm_api_key is None
+
+    # Mixed quotes ("value' → unchanged)
+    assert config.fallback_endpoint == "\"mixed_quote'"
+
+    # Strings with internal quotes ("internal\"quotes" → internal"quotes")
+    assert config.baml_llm_model == 'internal"quotes'

cognee/tests/unit/infrastructure/mock_embedding_engine.py

@@ -4,10 +4,7 @@ from typing import List
 from cognee.infrastructure.databases.vector.embeddings.LiteLLMEmbeddingEngine import (
     LiteLLMEmbeddingEngine,
 )
-from cognee.infrastructure.databases.vector.embeddings.embedding_rate_limiter import (
-    embedding_rate_limit_async,
-    embedding_sleep_and_retry_async,
-)
+from cognee.shared.rate_limiting import embedding_rate_limiter_context_manager
 
 
 class MockEmbeddingEngine(LiteLLMEmbeddingEngine):
@@ -34,8 +31,6 @@ class MockEmbeddingEngine(LiteLLMEmbeddingEngine):
         self.fail_every_n_requests = fail_every_n_requests
         self.add_delay = add_delay
 
-    @embedding_sleep_and_retry_async()
-    @embedding_rate_limit_async
     async def embed_text(self, text: List[str]) -> List[List[float]]:
         """
         Mock implementation that returns fixed embeddings and can
@@ -52,4 +47,5 @@ class MockEmbeddingEngine(LiteLLMEmbeddingEngine):
             raise Exception(f"Mock failure on request #{self.request_count}")
 
         # Return mock embeddings of the correct dimension
-        return [[0.1] * self.dimensions for _ in text]
+        async with embedding_rate_limiter_context_manager():
+            return [[0.1] * self.dimensions for _ in text]

cognee/tests/unit/infrastructure/test_embedding_rate_limiting_realistic.py

@@ -6,9 +6,6 @@ import logging
 from cognee.infrastructure.llm.config import (
     get_llm_config,
 )
-from cognee.infrastructure.databases.vector.embeddings.embedding_rate_limiter import (
-    EmbeddingRateLimiter,
-)
 from cognee.tests.unit.infrastructure.mock_embedding_engine import MockEmbeddingEngine
 
 # Configure logging
@@ -33,7 +30,6 @@ async def test_embedding_rate_limiting_realistic():
 
     # Clear the config and rate limiter caches to ensure our settings are applied
     get_llm_config.cache_clear()
-    EmbeddingRateLimiter.reset_instance()
 
     # Create a fresh config instance and verify settings
     config = get_llm_config()
@@ -170,7 +166,6 @@ async def test_with_mock_failures():
 
     # Clear caches
     get_llm_config.cache_clear()
-    EmbeddingRateLimiter.reset_instance()
 
     # Create a mock engine configured to fail every 3rd request
     engine = MockEmbeddingEngine()