FAISS and embedding support (#48)

also adds embeddings and an in memory docstore

tests/integration_tests/embeddings/__init__.py

@@ -0,0 +1 @@
+"""Test embedding integrations."""

tests/integration_tests/embeddings/test_openai.py

@@ -0,0 +1,19 @@
+"""Test openai embeddings."""
+from langchain.embeddings.openai import OpenAIEmbeddings
+
+
+def test_openai_embedding_documents() -> None:
+    """Test openai embeddings."""
+    documents = ["foo bar"]
+    embedding = OpenAIEmbeddings()
+    output = embedding.embed_documents(documents)
+    assert len(output) == 1
+    assert len(output[0]) == 2048
+
+
+def test_openai_embedding_query() -> None:
+    """Test openai embeddings."""
+    document = "foo bar"
+    embedding = OpenAIEmbeddings()
+    output = embedding.embed_query(document)
+    assert len(output) == 2048

tests/integration_tests/test_faiss.py

@@ -0,0 +1,47 @@
+"""Test FAISS functionality."""
+from typing import List
+
+import pytest
+
+from langchain.docstore.document import Document
+from langchain.docstore.in_memory import InMemoryDocstore
+from langchain.embeddings.base import Embeddings
+from langchain.faiss import FAISS
+
+
+class FakeEmbeddings(Embeddings):
+    """Fake embeddings functionality for testing."""
+
+    def embed_documents(self, texts: List[str]) -> List[List[float]]:
+        """Return simple embeddings."""
+        return [[i] * 10 for i in range(len(texts))]
+
+    def embed_query(self, text: str) -> List[float]:
+        """Return simple embeddings."""
+        return [0] * 10
+
+
+def test_faiss() -> None:
+    """Test end to end construction and search."""
+    texts = ["foo", "bar", "baz"]
+    docsearch = FAISS.from_texts(texts, FakeEmbeddings())
+    expected_docstore = InMemoryDocstore(
+        {
+            "0": Document(page_content="foo"),
+            "1": Document(page_content="bar"),
+            "2": Document(page_content="baz"),
+        }
+    )
+    assert docsearch.docstore.__dict__ == expected_docstore.__dict__
+    output = docsearch.similarity_search("foo", k=1)
+    assert output == [Document(page_content="foo")]
+
+
+def test_faiss_search_not_found() -> None:
+    """Test what happens when document is not found."""
+    texts = ["foo", "bar", "baz"]
+    docsearch = FAISS.from_texts(texts, FakeEmbeddings())
+    # Get rid of the docstore to purposefully induce errors.
+    docsearch.docstore = InMemoryDocstore({})
+    with pytest.raises(ValueError):
+        docsearch.similarity_search("foo")

tests/unit_tests/docstore/test_inmemory.py

@@ -0,0 +1,21 @@
+"""Test in memory docstore."""
+
+from langchain.docstore.document import Document
+from langchain.docstore.in_memory import InMemoryDocstore
+
+
+def test_document_found() -> None:
+    """Test document found."""
+    _dict = {"foo": Document(page_content="bar")}
+    docstore = InMemoryDocstore(_dict)
+    output = docstore.search("foo")
+    assert isinstance(output, Document)
+    assert output.page_content == "bar"
+
+
+def test_document_not_found() -> None:
+    """Test when document is not found."""
+    _dict = {"foo": Document(page_content="bar")}
+    docstore = InMemoryDocstore(_dict)
+    output = docstore.search("bar")
+    assert output == "ID bar not found."