cognee 0.5.0.dev1__py3-none-any.whl → 0.5.1.dev0__py3-none-any.whl

cognee/tests/integration/retrieval/test_graph_completion_retriever_context_extension.py

@@ -0,0 +1,226 @@
+import os
+import pytest
+import pathlib
+import pytest_asyncio
+from typing import Optional, Union
+import cognee
+
+from cognee.low_level import setup, DataPoint
+from cognee.tasks.storage import add_data_points
+from cognee.modules.graph.utils import resolve_edges_to_text
+from cognee.modules.retrieval.graph_completion_context_extension_retriever import (
+    GraphCompletionContextExtensionRetriever,
+)
+
+
+@pytest_asyncio.fixture
+async def setup_test_environment_simple():
+    """Set up a clean test environment with simple graph data."""
+    base_dir = pathlib.Path(__file__).parent.parent.parent.parent
+    system_directory_path = str(
+        base_dir / ".cognee_system/test_graph_completion_extension_context_simple"
+    )
+    data_directory_path = str(
+        base_dir / ".data_storage/test_graph_completion_extension_context_simple"
+    )
+
+    cognee.config.system_root_directory(system_directory_path)
+    cognee.config.data_root_directory(data_directory_path)
+
+    await cognee.prune.prune_data()
+    await cognee.prune.prune_system(metadata=True)
+    await setup()
+
+    class Company(DataPoint):
+        name: str
+
+    class Person(DataPoint):
+        name: str
+        works_for: Company
+
+    company1 = Company(name="Figma")
+    company2 = Company(name="Canva")
+    person1 = Person(name="Steve Rodger", works_for=company1)
+    person2 = Person(name="Ike Loma", works_for=company1)
+    person3 = Person(name="Jason Statham", works_for=company1)
+    person4 = Person(name="Mike Broski", works_for=company2)
+    person5 = Person(name="Christina Mayer", works_for=company2)
+
+    entities = [company1, company2, person1, person2, person3, person4, person5]
+
+    await add_data_points(entities)
+
+    yield
+
+    try:
+        await cognee.prune.prune_data()
+        await cognee.prune.prune_system(metadata=True)
+    except Exception:
+        pass
+
+
+@pytest_asyncio.fixture
+async def setup_test_environment_complex():
+    """Set up a clean test environment with complex graph data."""
+    base_dir = pathlib.Path(__file__).parent.parent.parent.parent
+    system_directory_path = str(
+        base_dir / ".cognee_system/test_graph_completion_extension_context_complex"
+    )
+    data_directory_path = str(
+        base_dir / ".data_storage/test_graph_completion_extension_context_complex"
+    )
+
+    cognee.config.system_root_directory(system_directory_path)
+    cognee.config.data_root_directory(data_directory_path)
+
+    await cognee.prune.prune_data()
+    await cognee.prune.prune_system(metadata=True)
+    await setup()
+
+    class Company(DataPoint):
+        name: str
+        metadata: dict = {"index_fields": ["name"]}
+
+    class Car(DataPoint):
+        brand: str
+        model: str
+        year: int
+
+    class Location(DataPoint):
+        country: str
+        city: str
+
+    class Home(DataPoint):
+        location: Location
+        rooms: int
+        sqm: int
+
+    class Person(DataPoint):
+        name: str
+        works_for: Company
+        owns: Optional[list[Union[Car, Home]]] = None
+
+    company1 = Company(name="Figma")
+    company2 = Company(name="Canva")
+
+    person1 = Person(name="Mike Rodger", works_for=company1)
+    person1.owns = [Car(brand="Toyota", model="Camry", year=2020)]
+
+    person2 = Person(name="Ike Loma", works_for=company1)
+    person2.owns = [
+        Car(brand="Tesla", model="Model S", year=2021),
+        Home(location=Location(country="USA", city="New York"), sqm=80, rooms=4),
+    ]
+
+    person3 = Person(name="Jason Statham", works_for=company1)
+
+    person4 = Person(name="Mike Broski", works_for=company2)
+    person4.owns = [Car(brand="Ford", model="Mustang", year=1978)]
+
+    person5 = Person(name="Christina Mayer", works_for=company2)
+    person5.owns = [Car(brand="Honda", model="Civic", year=2023)]
+
+    entities = [company1, company2, person1, person2, person3, person4, person5]
+
+    await add_data_points(entities)
+
+    yield
+
+    try:
+        await cognee.prune.prune_data()
+        await cognee.prune.prune_system(metadata=True)
+    except Exception:
+        pass
+
+
+@pytest_asyncio.fixture
+async def setup_test_environment_empty():
+    """Set up a clean test environment without graph data."""
+    base_dir = pathlib.Path(__file__).parent.parent.parent.parent
+    system_directory_path = str(
+        base_dir / ".cognee_system/test_get_graph_completion_extension_context_on_empty_graph"
+    )
+    data_directory_path = str(
+        base_dir / ".data_storage/test_get_graph_completion_extension_context_on_empty_graph"
+    )
+
+    cognee.config.system_root_directory(system_directory_path)
+    cognee.config.data_root_directory(data_directory_path)
+
+    await cognee.prune.prune_data()
+    await cognee.prune.prune_system(metadata=True)
+    await setup()
+
+    yield
+
+    try:
+        await cognee.prune.prune_data()
+        await cognee.prune.prune_system(metadata=True)
+    except Exception:
+        pass
+
+
+@pytest.mark.asyncio
+async def test_graph_completion_extension_context_simple(setup_test_environment_simple):
+    """Integration test: verify GraphCompletionContextExtensionRetriever can retrieve context (simple)."""
+    retriever = GraphCompletionContextExtensionRetriever()
+
+    context = await resolve_edges_to_text(await retriever.get_context("Who works at Canva?"))
+
+    assert "Mike Broski --[works_for]--> Canva" in context, "Failed to get Mike Broski"
+    assert "Christina Mayer --[works_for]--> Canva" in context, "Failed to get Christina Mayer"
+
+    answer = await retriever.get_completion("Who works at Canva?")
+
+    assert isinstance(answer, list), f"Expected list, got {type(answer).__name__}"
+    assert all(isinstance(item, str) and item.strip() for item in answer), (
+        "Answer must contain only non-empty strings"
+    )
+
+
+@pytest.mark.asyncio
+async def test_graph_completion_extension_context_complex(setup_test_environment_complex):
+    """Integration test: verify GraphCompletionContextExtensionRetriever can retrieve context (complex)."""
+    retriever = GraphCompletionContextExtensionRetriever(top_k=20)
+
+    context = await resolve_edges_to_text(
+        await retriever.get_context("Who works at Figma and drives Tesla?")
+    )
+
+    assert "Mike Rodger --[works_for]--> Figma" in context, "Failed to get Mike Rodger"
+    assert "Ike Loma --[works_for]--> Figma" in context, "Failed to get Ike Loma"
+    assert "Jason Statham --[works_for]--> Figma" in context, "Failed to get Jason Statham"
+
+    answer = await retriever.get_completion("Who works at Figma?")
+
+    assert isinstance(answer, list), f"Expected list, got {type(answer).__name__}"
+    assert all(isinstance(item, str) and item.strip() for item in answer), (
+        "Answer must contain only non-empty strings"
+    )
+
+
+@pytest.mark.asyncio
+async def test_get_graph_completion_extension_context_on_empty_graph(setup_test_environment_empty):
+    """Integration test: verify GraphCompletionContextExtensionRetriever handles empty graph correctly."""
+    retriever = GraphCompletionContextExtensionRetriever()
+
+    context = await retriever.get_context("Who works at Figma?")
+    assert context == [], "Context should be empty on an empty graph"
+
+    answer = await retriever.get_completion("Who works at Figma?")
+
+    assert isinstance(answer, list), f"Expected list, got {type(answer).__name__}"
+    assert all(isinstance(item, str) and item.strip() for item in answer), (
+        "Answer must contain only non-empty strings"
+    )
+
+
+@pytest.mark.asyncio
+async def test_graph_completion_extension_get_triplets_empty(setup_test_environment_empty):
+    """Integration test: verify GraphCompletionContextExtensionRetriever get_triplets handles empty graph."""
+    retriever = GraphCompletionContextExtensionRetriever()
+
+    triplets = await retriever.get_triplets("Who works at Figma?")
+
+    assert isinstance(triplets, list), "Triplets should be a list"
+    assert len(triplets) == 0, "Should return empty list on empty graph"

cognee/tests/integration/retrieval/test_graph_completion_retriever_cot.py

@@ -0,0 +1,218 @@
+import os
+import pytest
+import pathlib
+import pytest_asyncio
+from typing import Optional, Union
+import cognee
+
+from cognee.low_level import setup, DataPoint
+from cognee.modules.graph.utils import resolve_edges_to_text
+from cognee.tasks.storage import add_data_points
+from cognee.modules.retrieval.graph_completion_cot_retriever import GraphCompletionCotRetriever
+
+
+@pytest_asyncio.fixture
+async def setup_test_environment_simple():
+    """Set up a clean test environment with simple graph data."""
+    base_dir = pathlib.Path(__file__).parent.parent.parent.parent
+    system_directory_path = str(
+        base_dir / ".cognee_system/test_graph_completion_cot_context_simple"
+    )
+    data_directory_path = str(base_dir / ".data_storage/test_graph_completion_cot_context_simple")
+
+    cognee.config.system_root_directory(system_directory_path)
+    cognee.config.data_root_directory(data_directory_path)
+
+    await cognee.prune.prune_data()
+    await cognee.prune.prune_system(metadata=True)
+    await setup()
+
+    class Company(DataPoint):
+        name: str
+
+    class Person(DataPoint):
+        name: str
+        works_for: Company
+
+    company1 = Company(name="Figma")
+    company2 = Company(name="Canva")
+    person1 = Person(name="Steve Rodger", works_for=company1)
+    person2 = Person(name="Ike Loma", works_for=company1)
+    person3 = Person(name="Jason Statham", works_for=company1)
+    person4 = Person(name="Mike Broski", works_for=company2)
+    person5 = Person(name="Christina Mayer", works_for=company2)
+
+    entities = [company1, company2, person1, person2, person3, person4, person5]
+
+    await add_data_points(entities)
+
+    yield
+
+    try:
+        await cognee.prune.prune_data()
+        await cognee.prune.prune_system(metadata=True)
+    except Exception:
+        pass
+
+
+@pytest_asyncio.fixture
+async def setup_test_environment_complex():
+    """Set up a clean test environment with complex graph data."""
+    base_dir = pathlib.Path(__file__).parent.parent.parent.parent
+    system_directory_path = str(
+        base_dir / ".cognee_system/test_graph_completion_cot_context_complex"
+    )
+    data_directory_path = str(base_dir / ".data_storage/test_graph_completion_cot_context_complex")
+
+    cognee.config.system_root_directory(system_directory_path)
+    cognee.config.data_root_directory(data_directory_path)
+
+    await cognee.prune.prune_data()
+    await cognee.prune.prune_system(metadata=True)
+    await setup()
+
+    class Company(DataPoint):
+        name: str
+        metadata: dict = {"index_fields": ["name"]}
+
+    class Car(DataPoint):
+        brand: str
+        model: str
+        year: int
+
+    class Location(DataPoint):
+        country: str
+        city: str
+
+    class Home(DataPoint):
+        location: Location
+        rooms: int
+        sqm: int
+
+    class Person(DataPoint):
+        name: str
+        works_for: Company
+        owns: Optional[list[Union[Car, Home]]] = None
+
+    company1 = Company(name="Figma")
+    company2 = Company(name="Canva")
+
+    person1 = Person(name="Mike Rodger", works_for=company1)
+    person1.owns = [Car(brand="Toyota", model="Camry", year=2020)]
+
+    person2 = Person(name="Ike Loma", works_for=company1)
+    person2.owns = [
+        Car(brand="Tesla", model="Model S", year=2021),
+        Home(location=Location(country="USA", city="New York"), sqm=80, rooms=4),
+    ]
+
+    person3 = Person(name="Jason Statham", works_for=company1)
+
+    person4 = Person(name="Mike Broski", works_for=company2)
+    person4.owns = [Car(brand="Ford", model="Mustang", year=1978)]
+
+    person5 = Person(name="Christina Mayer", works_for=company2)
+    person5.owns = [Car(brand="Honda", model="Civic", year=2023)]
+
+    entities = [company1, company2, person1, person2, person3, person4, person5]
+
+    await add_data_points(entities)
+
+    yield
+
+    try:
+        await cognee.prune.prune_data()
+        await cognee.prune.prune_system(metadata=True)
+    except Exception:
+        pass
+
+
+@pytest_asyncio.fixture
+async def setup_test_environment_empty():
+    """Set up a clean test environment without graph data."""
+    base_dir = pathlib.Path(__file__).parent.parent.parent.parent
+    system_directory_path = str(
+        base_dir / ".cognee_system/test_get_graph_completion_cot_context_on_empty_graph"
+    )
+    data_directory_path = str(
+        base_dir / ".data_storage/test_get_graph_completion_cot_context_on_empty_graph"
+    )
+
+    cognee.config.system_root_directory(system_directory_path)
+    cognee.config.data_root_directory(data_directory_path)
+
+    await cognee.prune.prune_data()
+    await cognee.prune.prune_system(metadata=True)
+    await setup()
+
+    yield
+
+    try:
+        await cognee.prune.prune_data()
+        await cognee.prune.prune_system(metadata=True)
+    except Exception:
+        pass
+
+
+@pytest.mark.asyncio
+async def test_graph_completion_cot_context_simple(setup_test_environment_simple):
+    """Integration test: verify GraphCompletionCotRetriever can retrieve context (simple)."""
+    retriever = GraphCompletionCotRetriever()
+
+    context = await resolve_edges_to_text(await retriever.get_context("Who works at Canva?"))
+
+    assert "Mike Broski --[works_for]--> Canva" in context, "Failed to get Mike Broski"
+    assert "Christina Mayer --[works_for]--> Canva" in context, "Failed to get Christina Mayer"
+
+    answer = await retriever.get_completion("Who works at Canva?")
+
+    assert isinstance(answer, list), f"Expected list, got {type(answer).__name__}"
+    assert all(isinstance(item, str) and item.strip() for item in answer), (
+        "Answer must contain only non-empty strings"
+    )
+
+
+@pytest.mark.asyncio
+async def test_graph_completion_cot_context_complex(setup_test_environment_complex):
+    """Integration test: verify GraphCompletionCotRetriever can retrieve context (complex)."""
+    retriever = GraphCompletionCotRetriever(top_k=20)
+
+    context = await resolve_edges_to_text(await retriever.get_context("Who works at Figma?"))
+
+    assert "Mike Rodger --[works_for]--> Figma" in context, "Failed to get Mike Rodger"
+    assert "Ike Loma --[works_for]--> Figma" in context, "Failed to get Ike Loma"
+    assert "Jason Statham --[works_for]--> Figma" in context, "Failed to get Jason Statham"
+
+    answer = await retriever.get_completion("Who works at Figma?")
+
+    assert isinstance(answer, list), f"Expected list, got {type(answer).__name__}"
+    assert all(isinstance(item, str) and item.strip() for item in answer), (
+        "Answer must contain only non-empty strings"
+    )
+
+
+@pytest.mark.asyncio
+async def test_get_graph_completion_cot_context_on_empty_graph(setup_test_environment_empty):
+    """Integration test: verify GraphCompletionCotRetriever handles empty graph correctly."""
+    retriever = GraphCompletionCotRetriever()
+
+    context = await retriever.get_context("Who works at Figma?")
+    assert context == [], "Context should be empty on an empty graph"
+
+    answer = await retriever.get_completion("Who works at Figma?")
+
+    assert isinstance(answer, list), f"Expected list, got {type(answer).__name__}"
+    assert all(isinstance(item, str) and item.strip() for item in answer), (
+        "Answer must contain only non-empty strings"
+    )
+
+
+@pytest.mark.asyncio
+async def test_graph_completion_cot_get_triplets_empty(setup_test_environment_empty):
+    """Integration test: verify GraphCompletionCotRetriever get_triplets handles empty graph."""
+    retriever = GraphCompletionCotRetriever()
+
+    triplets = await retriever.get_triplets("Who works at Figma?")
+
+    assert isinstance(triplets, list), "Triplets should be a list"
+    assert len(triplets) == 0, "Should return empty list on empty graph"

cognee/tests/integration/retrieval/test_rag_completion_retriever.py

@@ -0,0 +1,254 @@
+import os
+from typing import List
+import pytest
+import pathlib
+import pytest_asyncio
+import cognee
+
+from cognee.low_level import setup
+from cognee.tasks.storage import add_data_points
+from cognee.infrastructure.databases.vector import get_vector_engine
+from cognee.modules.chunking.models import DocumentChunk
+from cognee.modules.data.processing.document_types import TextDocument
+from cognee.modules.retrieval.exceptions.exceptions import NoDataError
+from cognee.modules.retrieval.completion_retriever import CompletionRetriever
+from cognee.infrastructure.engine import DataPoint
+from cognee.modules.data.processing.document_types import Document
+from cognee.modules.engine.models import Entity
+
+
+class DocumentChunkWithEntities(DataPoint):
+    text: str
+    chunk_size: int
+    chunk_index: int
+    cut_type: str
+    is_part_of: Document
+    contains: List[Entity] = None
+
+    metadata: dict = {"index_fields": ["text"]}
+
+
+@pytest_asyncio.fixture
+async def setup_test_environment_with_chunks_simple():
+    """Set up a clean test environment with simple chunks."""
+    base_dir = pathlib.Path(__file__).parent.parent.parent.parent
+    system_directory_path = str(base_dir / ".cognee_system/test_rag_completion_context_simple")
+    data_directory_path = str(base_dir / ".data_storage/test_rag_completion_context_simple")
+
+    cognee.config.system_root_directory(system_directory_path)
+    cognee.config.data_root_directory(data_directory_path)
+
+    await cognee.prune.prune_data()
+    await cognee.prune.prune_system(metadata=True)
+    await setup()
+
+    document = TextDocument(
+        name="Steve Rodger's career",
+        raw_data_location="somewhere",
+        external_metadata="",
+        mime_type="text/plain",
+    )
+
+    chunk1 = DocumentChunk(
+        text="Steve Rodger",
+        chunk_size=2,
+        chunk_index=0,
+        cut_type="sentence_end",
+        is_part_of=document,
+        contains=[],
+    )
+    chunk2 = DocumentChunk(
+        text="Mike Broski",
+        chunk_size=2,
+        chunk_index=1,
+        cut_type="sentence_end",
+        is_part_of=document,
+        contains=[],
+    )
+    chunk3 = DocumentChunk(
+        text="Christina Mayer",
+        chunk_size=2,
+        chunk_index=2,
+        cut_type="sentence_end",
+        is_part_of=document,
+        contains=[],
+    )
+
+    entities = [chunk1, chunk2, chunk3]
+
+    await add_data_points(entities)
+
+    yield
+
+    try:
+        await cognee.prune.prune_data()
+        await cognee.prune.prune_system(metadata=True)
+    except Exception:
+        pass
+
+
+@pytest_asyncio.fixture
+async def setup_test_environment_with_chunks_complex():
+    """Set up a clean test environment with complex chunks."""
+    base_dir = pathlib.Path(__file__).parent.parent.parent.parent
+    system_directory_path = str(base_dir / ".cognee_system/test_rag_completion_context_complex")
+    data_directory_path = str(base_dir / ".data_storage/test_rag_completion_context_complex")
+
+    cognee.config.system_root_directory(system_directory_path)
+    cognee.config.data_root_directory(data_directory_path)
+
+    await cognee.prune.prune_data()
+    await cognee.prune.prune_system(metadata=True)
+    await setup()
+
+    document1 = TextDocument(
+        name="Employee List",
+        raw_data_location="somewhere",
+        external_metadata="",
+        mime_type="text/plain",
+    )
+
+    document2 = TextDocument(
+        name="Car List",
+        raw_data_location="somewhere",
+        external_metadata="",
+        mime_type="text/plain",
+    )
+
+    chunk1 = DocumentChunk(
+        text="Steve Rodger",
+        chunk_size=2,
+        chunk_index=0,
+        cut_type="sentence_end",
+        is_part_of=document1,
+        contains=[],
+    )
+    chunk2 = DocumentChunk(
+        text="Mike Broski",
+        chunk_size=2,
+        chunk_index=1,
+        cut_type="sentence_end",
+        is_part_of=document1,
+        contains=[],
+    )
+    chunk3 = DocumentChunk(
+        text="Christina Mayer",
+        chunk_size=2,
+        chunk_index=2,
+        cut_type="sentence_end",
+        is_part_of=document1,
+        contains=[],
+    )
+
+    chunk4 = DocumentChunk(
+        text="Range Rover",
+        chunk_size=2,
+        chunk_index=0,
+        cut_type="sentence_end",
+        is_part_of=document2,
+        contains=[],
+    )
+    chunk5 = DocumentChunk(
+        text="Hyundai",
+        chunk_size=2,
+        chunk_index=1,
+        cut_type="sentence_end",
+        is_part_of=document2,
+        contains=[],
+    )
+    chunk6 = DocumentChunk(
+        text="Chrysler",
+        chunk_size=2,
+        chunk_index=2,
+        cut_type="sentence_end",
+        is_part_of=document2,
+        contains=[],
+    )
+
+    entities = [chunk1, chunk2, chunk3, chunk4, chunk5, chunk6]
+
+    await add_data_points(entities)
+
+    yield
+
+    try:
+        await cognee.prune.prune_data()
+        await cognee.prune.prune_system(metadata=True)
+    except Exception:
+        pass
+
+
+@pytest_asyncio.fixture
+async def setup_test_environment_empty():
+    """Set up a clean test environment without chunks."""
+    base_dir = pathlib.Path(__file__).parent.parent.parent.parent
+    system_directory_path = str(
+        base_dir / ".cognee_system/test_get_rag_completion_context_on_empty_graph"
+    )
+    data_directory_path = str(
+        base_dir / ".data_storage/test_get_rag_completion_context_on_empty_graph"
+    )
+
+    cognee.config.system_root_directory(system_directory_path)
+    cognee.config.data_root_directory(data_directory_path)
+
+    await cognee.prune.prune_data()
+    await cognee.prune.prune_system(metadata=True)
+
+    yield
+
+    try:
+        await cognee.prune.prune_data()
+        await cognee.prune.prune_system(metadata=True)
+    except Exception:
+        pass
+
+
+@pytest.mark.asyncio
+async def test_rag_completion_context_simple(setup_test_environment_with_chunks_simple):
+    """Integration test: verify CompletionRetriever can retrieve context (simple)."""
+    retriever = CompletionRetriever()
+
+    context = await retriever.get_context("Mike")
+
+    assert isinstance(context, str), "Context should be a string"
+    assert "Mike Broski" in context, "Failed to get Mike Broski"
+
+
+@pytest.mark.asyncio
+async def test_rag_completion_context_multiple_chunks(setup_test_environment_with_chunks_simple):
+    """Integration test: verify CompletionRetriever can retrieve context from multiple chunks."""
+    retriever = CompletionRetriever()
+
+    context = await retriever.get_context("Steve")
+
+    assert isinstance(context, str), "Context should be a string"
+    assert "Steve Rodger" in context, "Failed to get Steve Rodger"
+
+
+@pytest.mark.asyncio
+async def test_rag_completion_context_complex(setup_test_environment_with_chunks_complex):
+    """Integration test: verify CompletionRetriever can retrieve context (complex)."""
+    # TODO: top_k doesn't affect the output, it should be fixed.
+    retriever = CompletionRetriever(top_k=20)
+
+    context = await retriever.get_context("Christina")
+
+    assert context[0:15] == "Christina Mayer", "Failed to get Christina Mayer"
+
+
+@pytest.mark.asyncio
+async def test_get_rag_completion_context_on_empty_graph(setup_test_environment_empty):
+    """Integration test: verify CompletionRetriever handles empty graph correctly."""
+    retriever = CompletionRetriever()
+
+    with pytest.raises(NoDataError):
+        await retriever.get_context("Christina Mayer")
+
+    vector_engine = get_vector_engine()
+    await vector_engine.create_collection(
+        "DocumentChunk_text", payload_schema=DocumentChunkWithEntities
+    )
+
+    context = await retriever.get_context("Christina Mayer")
+    assert context == "", "Returned context should be empty on an empty graph"