Add ruff rule for Error Messages (EM) (#646)

Author: cbornet

examples/evaluation/tru_shared.py

@@ -129,7 +129,8 @@ def get_recorder(
             feedbacks=feedbacks,
             feedback_mode=feedback_mode,
         )
-    raise ValueError(f"Unknown framework: {framework} specified for get_recorder()")
+    msg = f"Unknown framework: {framework} specified for get_recorder()"
+    raise ValueError(msg)
 
 
 def get_azure_chat_model(
@@ -151,7 +152,8 @@ def get_azure_chat_model(
             model_version=model_version,
             temperature=temperature,
         )
-    raise ValueError(f"Unknown framework: {framework} specified for getChatModel()")
+    msg = f"Unknown framework: {framework} specified for getChatModel()"
+    raise ValueError(msg)
 
 
 def get_azure_embeddings_model(framework: Framework):
@@ -167,9 +169,8 @@ def get_azure_embeddings_model(framework: Framework):
             api_version="2023-05-15",
             temperature=temperature,
         )
-    raise ValueError(
-        f"Unknown framework: {framework} specified for getEmbeddingsModel()"
-    )
+    msg = f"Unknown framework: {framework} specified for getEmbeddingsModel()"
+    raise ValueError(msg)
 
 
 def get_astra_vector_store(framework: Framework, collection_name: str):
@@ -187,9 +188,8 @@ def get_astra_vector_store(framework: Framework, collection_name: str):
             token=os.getenv("ASTRA_DB_APPLICATION_TOKEN"),
             embedding_dimension=1536,
         )
-    raise ValueError(
-        f"Unknown framework: {framework} specified for get_astra_vector_store()"
-    )
+    msg = f"Unknown framework: {framework} specified for get_astra_vector_store()"
+    raise ValueError(msg)
 
 
 def execute_query(framework: Framework, pipeline, query) -> None:
@@ -198,9 +198,8 @@ def execute_query(framework: Framework, pipeline, query) -> None:
     elif framework == Framework.LLAMA_INDEX:
         pipeline.query(query)
     else:
-        raise ValueError(
-            f"Unknown framework: {framework} specified for execute_query()"
-        )
+        msg = f"Unknown framework: {framework} specified for execute_query()"
+        raise ValueError(msg)
 
 
 # runs the pipeline across all queries in all known datasets

examples/notebooks/advancedRAG.ipynb

@@ -162,7 +162,8 @@
     "if uploaded:\n",
     "    SAMPLEDATA = uploaded\n",
     "else:\n",
-    "    raise ValueError(\"Cannot proceed without Sample Data. Please re-run the cell.\")\n",
+    "    msg = \"Cannot proceed without Sample Data. Please re-run the cell.\"\n",
+    "    raise ValueError(msg)\n",
     "\n",
     "print(\"Please make sure to change your queries to match the contents of your file!\")"
    ]

examples/notebooks/conftest.py

@@ -8,10 +8,12 @@
 
 def get_required_env(name) -> str:
     if name not in os.environ:
-        raise ValueError(f"Missing required environment variable: {name}")
+        msg = f"Missing required environment variable: {name}"
+        raise ValueError(msg)
     value = os.environ[name]
     if not value:
-        raise ValueError(f"Empty required environment variable: {name}")
+        msg = f"Empty required environment variable: {name}"
+        raise ValueError(msg)
     return value
 
 

examples/notebooks/langchain_evaluation.ipynb

@@ -193,7 +193,8 @@
     "if uploaded:\n",
     "    SAMPLEDATA = uploaded\n",
     "else:\n",
-    "    raise ValueError(\"Cannot proceed without Sample Data. Please re-run the cell.\")\n",
+    "    msg = \"Cannot proceed without Sample Data. Please re-run the cell.\"\n",
+    "    raise ValueError(msg)\n",
     "\n",
     "print(\"Please make sure to change your queries to match the contents of your file!\")"
    ]

libs/colbert/ragstack_colbert/cassandra_database.py

@@ -45,10 +45,11 @@ class CassandraDatabase(BaseDatabase):
     _table: ClusteredMetadataVectorCassandraTable
 
     def __new__(cls) -> Self:  # noqa: D102
-        raise ValueError(
+        msg = (
             "This class cannot be instantiated directly. "
             "Please use the `from_astra()` or `from_session()` class methods."
         )
+        raise ValueError(msg)
 
     @classmethod
     def from_astra(
@@ -173,10 +174,11 @@ def add_chunks(self, chunks: list[Chunk]) -> list[tuple[str, int]]:
             success_chunks.append((doc_id, chunk_id))
 
         if len(failed_chunks) > 0:
-            raise CassandraDatabaseError(
+            msg = (
                 f"add failed for these chunks: {failed_chunks}. "
                 f"See error logs for more info."
             )
+            raise CassandraDatabaseError(msg)
 
         return success_chunks
 
@@ -273,10 +275,11 @@ async def aadd_chunks(
                 failed_chunks.append((doc_id, chunk_id))
 
         if len(failed_chunks) > 0:
-            raise CassandraDatabaseError(
+            msg = (
                 f"add failed for these chunks: {failed_chunks}. "
                 f"See error logs for more info."
             )
+            raise CassandraDatabaseError(msg)
 
         return outputs
 
@@ -292,8 +295,9 @@ def delete_chunks(self, doc_ids: list[str]) -> bool:
                 failed_docs.append(doc_id)
 
         if len(failed_docs) > 0:
+            msg = "delete failed for these docs: %s. See error logs for more info."
             raise CassandraDatabaseError(
-                "delete failed for these docs: %s. See error logs for more info.",
+                msg,
                 failed_docs,
             )
 
@@ -340,10 +344,11 @@ async def adelete_chunks(
                     failed_docs.append(doc_id)
 
         if len(failed_docs) > 0:
-            raise CassandraDatabaseError(
+            msg = (
                 f"delete failed for these docs: {failed_docs}. "
                 f"See error logs for more info."
             )
+            raise CassandraDatabaseError(msg)
 
         return success
 
@@ -379,9 +384,8 @@ async def get_chunk_data(
         row = await self._table.aget(partition_id=doc_id, row_id=row_id)
 
         if row is None:
-            raise CassandraDatabaseError(
-                f"no chunk found for doc_id: {doc_id} chunk_id: {chunk_id}"
-            )
+            msg = f"no chunk found for doc_id: {doc_id} chunk_id: {chunk_id}"
+            raise CassandraDatabaseError(msg)
 
         if include_embedding is True:
             embedded_chunk = await self.get_chunk_embedding(

libs/colbert/ragstack_colbert/colbert_vector_store.py

@@ -46,9 +46,8 @@ def __init__(
 
     def _validate_embedding_model(self) -> BaseEmbeddingModel:
         if self._embedding_model is None:
-            raise AttributeError(
-                "To use this method, `embedding_model` must be set on class creation."
-            )
+            msg = "To use this method, `embedding_model` must be set on class creation."
+            raise AttributeError(msg)
         return self._embedding_model
 
     def _build_chunks(
@@ -60,7 +59,8 @@ def _build_chunks(
         embedding_model = self._validate_embedding_model()
 
         if metadatas is not None and len(texts) != len(metadatas):
-            raise ValueError("Length of texts and metadatas must match.")
+            msg = "Length of texts and metadatas must match."
+            raise ValueError(msg)
 
         if doc_id is None:
             doc_id = str(uuid.uuid4())

libs/e2e-tests/e2e_tests/conftest.py

@@ -55,7 +55,8 @@ def get_required_env(name) -> str:
 
 vector_database_type = os.environ.get("VECTOR_DATABASE_TYPE", "astradb")
 if vector_database_type not in ["astradb", "local-cassandra"]:
-    raise ValueError(f"Invalid VECTOR_DATABASE_TYPE: {vector_database_type}")
+    msg = f"Invalid VECTOR_DATABASE_TYPE: {vector_database_type}"
+    raise ValueError(msg)
 
 is_astra = vector_database_type == "astradb"
 
@@ -67,7 +68,8 @@ def get_vector_store_handler(
         return AstraDBVectorStoreHandler(implementation)
     if vector_database_type == "local-cassandra":
         return CassandraVectorStoreHandler(implementation)
-    raise ValueError("Invalid vector store implementation")
+    msg = "Invalid vector store implementation"
+    raise ValueError(msg)
 
 
 failed_report_lines = []

libs/e2e-tests/e2e_tests/langchain/test_compatibility_rag.py

@@ -363,7 +363,8 @@ def _run_test(
             vector_store=vector_store, config=resolved_llm["nemo_config"]
         )
     else:
-        raise ValueError(f"Unknown test case: {test_case}")
+        msg = f"Unknown test case: {test_case}"
+        raise ValueError(msg)
 
 
 @pytest.fixture()

libs/knowledge-store/ragstack_knowledge_store/_mmr_helper.py

@@ -131,11 +131,12 @@ def _pop_candidate(self, candidate_id: str) -> NDArray[np.float32]:
         """
         # Get the embedding for the id.
         index = self.candidate_id_to_index.pop(candidate_id)
-        if not self.candidates[index].id == candidate_id:
-            raise ValueError(
+        if self.candidates[index].id != candidate_id:
+            msg = (
                 "ID in self.candidate_id_to_index doesn't match the ID of the "
                 "corresponding index in self.candidates"
             )
+            raise ValueError(msg)
         embedding: NDArray[np.float32] = self.candidate_embeddings[index].copy()
 
         # Swap that index with the last index in the candidates and

libs/knowledge-store/ragstack_knowledge_store/_utils.py

@@ -16,7 +16,8 @@
     # This is equivalent to `itertools.batched`, but that is only available in 3.12
     def batched(iterable: Iterable[T], n: int) -> Iterator[tuple[T, ...]]:
         if n < 1:
-            raise ValueError("n must be at least one")
+            msg = "n must be at least one"
+            raise ValueError(msg)
         it = iter(iterable)
         while batch := tuple(islice(it, n)):
             yield batch