chore(deps): update python-nonmajor

pyproject.toml

@@ -41,9 +41,9 @@ Changelog = "https://github.com/googleapis/langchain-google-cloud-sql-pg-python/
 test = [
     "black[jupyter]==25.1.0",
     "isort==6.0.0",
-    "mypy==1.11.2",
+    "mypy==1.13.0",
     "pytest-asyncio==0.24.0",
-    "pytest==8.3.3",
+    "pytest==8.3.4",
     "pytest-cov==6.0.0"
 ]
 

requirements.txt

@@ -1,5 +1,5 @@
-cloud-sql-python-connector[asyncpg]==1.12.1
-langchain-core==0.3.8
+cloud-sql-python-connector[asyncpg]==1.14.0
+langchain-core==0.3.22
 numpy==1.26.4
-pgvector==0.3.4
-SQLAlchemy[asyncio]==2.0.35
+pgvector==0.3.6
+SQLAlchemy[asyncio]==2.0.36

samples/index_tuning_sample/requirements.txt

@@ -1,3 +1,3 @@
 langchain-community==0.3.16
-langchain-google-cloud-sql-pg==0.10.0
+langchain-google-cloud-sql-pg==0.11.1
 langchain-google-vertexai==2.0.12

samples/langchain_on_vertexai/requirements.txt

@@ -1,5 +1,5 @@
 google-cloud-aiplatform[reasoningengine,langchain]==1.79.0
 google-cloud-resource-manager==1.14.0
 langchain-community==0.3.16
-langchain-google-cloud-sql-pg==0.10.0
+langchain-google-cloud-sql-pg==0.11.1
 langchain-google-vertexai==2.0.12

samples/requirements.txt

@@ -1,5 +1,5 @@
 google-cloud-aiplatform[reasoningengine,langchain]==1.79.0
 google-cloud-resource-manager==1.14.0
 langchain-community==0.3.16
-langchain-google-cloud-sql-pg==0.10.0
+langchain-google-cloud-sql-pg==0.11.1
 langchain-google-vertexai==2.0.12

tests/test_async_vectorstore_search.py

@@ -156,7 +156,9 @@ async def test_similarity_search_with_relevance_scores_threshold_cosine(self, vs
         results = await vs.asimilarity_search_with_relevance_scores(
             "foo", **score_threshold
         )
-        assert len(results) == 4
+        # Note: Since tests use FakeEmbeddings which are non-normalized vectors, results might have scores beyond the range [0,1].
+        # For a normalized embedding service, a threshold of zero will yield all matched documents.
+        assert len(results) == 2
 
         score_threshold = {"score_threshold": 0.02}
         results = await vs.asimilarity_search_with_relevance_scores(

tests/test_vectorstore_search.py

@@ -170,7 +170,9 @@ async def test_similarity_search_with_relevance_scores_threshold_cosine(self, vs
         results = await vs.asimilarity_search_with_relevance_scores(
             "foo", **score_threshold
         )
-        assert len(results) == 4
+        # Note: Since tests use FakeEmbeddings which are non-normalized vectors, results might have scores beyond the range [0,1].
+        # For a normalized embedding service, a threshold of zero will yield all matched documents.
+        assert len(results) == 2
 
         score_threshold = {"score_threshold": 0.02}
         results = await vs.asimilarity_search_with_relevance_scores(