test_incremental_delete_with_batch_size() — langchain Function Reference

Architecture documentation for the test_incremental_delete_with_batch_size() function in test_indexing.py from the langchain codebase.

Function python CoreAbstractions Serialization

Entity Profile

CoreAbstractions→ Serialization→ test_incremental_delete_with_batch_size() — langchain Function Reference

Dependency Diagram

graph TD
  2dca10f8_2f81_3b49_69dc_2dcdd91aa252["test_incremental_delete_with_batch_size()"]
  576ad89d_c8dc_eddf_9cd2_c8ae0e7c9978["test_indexing.py"]
  2dca10f8_2f81_3b49_69dc_2dcdd91aa252 -->|defined in| 576ad89d_c8dc_eddf_9cd2_c8ae0e7c9978
  style 2dca10f8_2f81_3b49_69dc_2dcdd91aa252 fill:#6366f1,stroke:#818cf8,color:#fff

Relationship Graph

Source Code

libs/core/tests/unit_tests/indexing/test_indexing.py lines 1647–1841

def test_incremental_delete_with_batch_size(
    record_manager: InMemoryRecordManager, vector_store: InMemoryVectorStore
) -> None:
    """Test indexing with incremental deletion strategy and batch size."""
    loader = ToyLoader(
        documents=[
            Document(
                page_content="1",
                metadata={"source": "1"},
            ),
            Document(
                page_content="2",
                metadata={"source": "2"},
            ),
            Document(
                page_content="3",
                metadata={"source": "3"},
            ),
            Document(
                page_content="4",
                metadata={"source": "4"},
            ),
        ]
    )

    with patch.object(
        record_manager,
        "get_time",
        return_value=datetime(2021, 1, 1, tzinfo=timezone.utc).timestamp(),
    ):
        assert index(
            loader,
            record_manager,
            vector_store,
            cleanup="incremental",
            source_id_key="source",
            batch_size=3,
            key_encoder="sha256",
        ) == {
            "num_added": 4,
            "num_deleted": 0,
            "num_skipped": 0,
            "num_updated": 0,
        }

    doc_texts = {
        # Ignoring type since doc should be in the store and not a None
        vector_store.get_by_ids([uid])[0].page_content
        for uid in vector_store.store
    }
    assert doc_texts == {"1", "2", "3", "4"}

    # Attempt to index again verify that nothing changes
    with patch.object(
        record_manager,
        "get_time",
        return_value=datetime(2021, 1, 2, tzinfo=timezone.utc).timestamp(),
    ):
        assert index(
            loader,
            record_manager,
            vector_store,
            cleanup="incremental",
            source_id_key="source",
            batch_size=3,
            key_encoder="sha256",
        ) == {
            "num_added": 0,
            "num_deleted": 0,
            "num_skipped": 4,
            "num_updated": 0,
        }

    doc_texts = {
        # Ignoring type since doc should be in the store and not a None
        vector_store.get_by_ids([uid])[0].page_content
        for uid in vector_store.store
    }
    assert doc_texts == {"1", "2", "3", "4"}

    # Attempt to index again verify that nothing changes