From 82ea6001f26cb11bc9515e079df20cbc036f24ac Mon Sep 17 00:00:00 2001
From: Zouhair Elhadi <zouhair11elhadi@gmail.com>
Date: Mon, 8 May 2023 23:15:32 +0100
Subject: [PATCH] update class Weaviate (weaviate.py): add an attribute
 (by_text) to check if we need to use query by text or by vector

---
 langchain/vectorstores/base.py     |  6 ++++++
 langchain/vectorstores/weaviate.py | 11 +++++++++++
 2 files changed, 17 insertions(+)

diff --git a/langchain/vectorstores/base.py b/langchain/vectorstores/base.py
index 53c9fa0cd8cb6..8dc2b85ddd138 100644
--- a/langchain/vectorstores/base.py
+++ b/langchain/vectorstores/base.py
@@ -107,6 +107,12 @@ def similarity_search(
     ) -> List[Document]:
         """Return docs most similar to query."""
 
+    @abstractmethod
+    def similarity_search_by_text(
+            self, query: str, k: int = 4, **kwargs: Any
+    ) -> List[Document]:
+        """Return docs most similar to query."""
+
     def similarity_search_with_relevance_scores(
         self,
         query: str,
diff --git a/langchain/vectorstores/weaviate.py b/langchain/vectorstores/weaviate.py
index 78d8beec2e905..ae90cb0f77461 100644
--- a/langchain/vectorstores/weaviate.py
+++ b/langchain/vectorstores/weaviate.py
@@ -83,6 +83,7 @@ def __init__(
         client: Any,
         index_name: str,
         text_key: str,
+        by_text: bool = True,
         embedding: Optional[Embeddings] = None,
         attributes: Optional[List[str]] = None,
         relevance_score_fn: Optional[
@@ -105,6 +106,7 @@ def __init__(
         self._index_name = index_name
         self._embedding = embedding
         self._text_key = text_key
+        self._by_text = by_text
         self._query_attrs = [self._text_key]
         self._relevance_score_fn = relevance_score_fn
         if attributes is not None:
@@ -154,6 +156,15 @@ def json_serializable(value: Any) -> Any:
         return ids
 
     def similarity_search(
+            self, query: str, k: int = 4, **kwargs: Any
+    ) -> List[Document]:
+        if self._by_text:
+            return self.similarity_search_by_text(query, k, **kwargs)
+        else:
+            embd = self._embedding.embed_query(query)
+            return self.similarity_search_by_vector(embd, k, **kwargs)
+
+    def similarity_search_by_text(
         self, query: str, k: int = 4, **kwargs: Any
     ) -> List[Document]:
         """Return docs most similar to query.