Add ruff rules for builtins (#565)

Author: cbornet

examples/notebooks/FLARE.ipynb

@@ -181,7 +181,7 @@
     "# Retrieve the text of a short story that will be indexed in the vector store\n",
     "# ruff: noqa: E501\n",
     "! curl https://raw.githubusercontent.com/CassioML/cassio-website/main/docs/frameworks/langchain/texts/amontillado.txt --output amontillado.txt\n",
-    "input = \"amontillado.txt\""
+    "text = \"amontillado.txt\""
    ]
   },
   {
@@ -193,7 +193,7 @@
     "from langchain.document_loaders import TextLoader\n",
     "\n",
     "# Load your input and split it into documents\n",
-    "loader = TextLoader(input)\n",
+    "loader = TextLoader(text)\n",
     "documents = loader.load_and_split()"
    ]
   },

libs/knowledge-graph/ragstack_knowledge_graph/render.py

@@ -20,8 +20,7 @@ def print_graph_documents(
         for relation in doc.relationships:
             source = relation.source
             target = relation.target
-            type = relation.type
-            print(f"{_node_label(source)} -> {_node_label(target)}: {type}")
+            print(f"{_node_label(source)} -> {_node_label(target)}: {relation.type}")
 
 
 def render_graph_documents(

libs/knowledge-graph/ragstack_knowledge_graph/traverse.py

@@ -15,7 +15,7 @@ class _Node(NamedTuple):
 class Node(_Node):
     __slots__ = ()
 
-    def __new__(cls, name, type, properties=None):
+    def __new__(cls, name, type, properties=None):  # noqa: A002
         if properties is None:
             properties = {}
         return super().__new__(cls, name, type, properties)

libs/knowledge-store/ragstack_knowledge_store/_mmr_helper.py

@@ -116,15 +116,15 @@ def _already_selected_embeddings(self) -> np.ndarray:
         selected = len(self.selected_ids)
         return np.vsplit(self.selected_embeddings, [selected])[0]
 
-    def _pop_candidate(self, id: str) -> np.ndarray:
+    def _pop_candidate(self, candidate_id: str) -> np.ndarray:
         """Pop the candidate with the given ID.
 
         Returns:
             The embedding of the candidate.
         """
         # Get the embedding for the id.
-        index = self.candidate_id_to_index.pop(id)
-        assert self.candidates[index].id == id
+        index = self.candidate_id_to_index.pop(candidate_id)
+        assert self.candidates[index].id == candidate_id
         embedding = self.candidate_embeddings[index].copy()
 
         # Swap that index with the last index in the candidates and
@@ -199,10 +199,10 @@ def add_candidates(self, candidates: Dict[str, List[float]]):
         # And add them to the
         new_embeddings = np.ndarray((len(include_ids), self.dimensions))
         offset = self.candidate_embeddings.shape[0]
-        for index, id in enumerate(include_ids):
-            if id in include_ids:
-                self.candidate_id_to_index[id] = offset + index
-                embedding = candidates[id]
+        for index, candidate_id in enumerate(include_ids):
+            if candidate_id in include_ids:
+                self.candidate_id_to_index[candidate_id] = offset + index
+                embedding = candidates[candidate_id]
                 new_embeddings[index] = embedding
 
         # Compute the similarity to the query.
@@ -213,12 +213,12 @@ def add_candidates(self, candidates: Dict[str, List[float]]):
         redundancy = cosine_similarity(
             new_embeddings, self._already_selected_embeddings()
         )
-        for index, id in enumerate(include_ids):
+        for index, candidate_id in enumerate(include_ids):
             max_redundancy = 0.0
             if redundancy.shape[0] > 0:
                 max_redundancy = redundancy[index].max()
             candidate = _Candidate(
-                id=id,
+                id=candidate_id,
                 weighted_similarity=self.lambda_mult * similarity[index][0],
                 weighted_redundancy=self.lambda_mult_complement * max_redundancy,
             )

libs/knowledge-store/ragstack_knowledge_store/graph_store.py

@@ -355,12 +355,14 @@ def add_nodes(rows):
                 for row in rows:
                     results[row.content_id] = _row_to_node(row)
 
-            for id in ids:
-                if id not in results:
-                    results[id] = None
-                    cq.execute(self._query_by_id, parameters=(id,), callback=add_nodes)
+            for node_id in ids:
+                if node_id not in results:
+                    results[node_id] = None
+                    cq.execute(
+                        self._query_by_id, parameters=(node_id,), callback=add_nodes
+                    )
 
-        return [results[id] for id in ids]
+        return [results[node_id] for node_id in ids]
 
     def mmr_traversal_search(
         self,
@@ -414,7 +416,7 @@ def mmr_traversal_search(
         helper.add_candidates({row.content_id: row.text_embedding for row in fetched})
 
         # Select the best item, K times.
-        depths = {id: 0 for id in helper.candidate_ids()}
+        depths = {candidate_id: 0 for candidate_id in helper.candidate_ids()}
         visited_tags = set()
         for _ in range(k):
             selected_id = helper.pop_best()
@@ -549,10 +551,10 @@ def visit_targets(d: int, targets: Sequence[NamedTuple]):
                         new_nodes_at_next_depth.add(content_id)
 
                 if new_nodes_at_next_depth:
-                    for id in new_nodes_at_next_depth:
+                    for node_id in new_nodes_at_next_depth:
                         cq.execute(
                             self._query_ids_and_link_to_tags_by_id,
-                            parameters=(id,),
+                            parameters=(node_id,),
                             callback=lambda rows, d=d: visit_nodes(d + 1, rows),
                         )
 

libs/langchain/ragstack_langchain/graph_store/extractors/gliner_link_extractor.py

@@ -39,7 +39,7 @@ def __init__(
         self._kind = kind
         self._extract_kwargs = extract_kwargs or {}
 
-    def extract_one(self, input: GLiNERInput) -> Set[Link]:
+    def extract_one(self, input: GLiNERInput) -> Set[Link]:  # noqa: A002
         return next(self.extract_many([input]))
 
     def extract_many(

libs/langchain/ragstack_langchain/graph_store/extractors/hierarchy_link_extractor.py

@@ -39,7 +39,7 @@ def as_document_extractor(
 
     def extract_one(
         self,
-        input: HierarchyInput,
+        input: HierarchyInput,  # noqa: A002
     ) -> Set[Link]:
         this_path = "/".join(input)
         parent_path = None

libs/langchain/ragstack_langchain/graph_store/extractors/html_link_extractor.py

@@ -102,7 +102,7 @@ def as_document_extractor(
 
     def extract_one(
         self,
-        input: HtmlInput,
+        input: HtmlInput,  # noqa: A002
     ) -> Set[Link]:
         content = input.content
         if isinstance(content, str):

libs/langchain/ragstack_langchain/graph_store/extractors/keybert_link_extractor.py

@@ -38,7 +38,7 @@ def __init__(
         self._kind = kind
         self._extract_keywords_kwargs = extract_keywords_kwargs or {}
 
-    def extract_one(self, input: KeybertInput) -> Set[Link]:
+    def extract_one(self, input: KeybertInput) -> Set[Link]:  # noqa: A002
         keywords = self._kw_model.extract_keywords(
             input if isinstance(input, str) else input.page_content,
             **self._extract_keywords_kwargs,

libs/langchain/ragstack_langchain/graph_store/extractors/link_extractor.py

@@ -15,7 +15,7 @@ class LinkExtractor(ABC, Generic[InputT]):
     """Interface for extracting links (incoming, outgoing, bidirectional)."""
 
     @abstractmethod
-    def extract_one(self, input: InputT) -> set[Link]:
+    def extract_one(self, input: InputT) -> set[Link]:  # noqa: A002
         """Add edges from each `input` to the corresponding documents.
 
         Args:
@@ -34,5 +34,4 @@ def extract_many(self, inputs: Iterable[InputT]) -> Iterable[Set[Link]]:
         Returns:
             Iterable over the set of links extracted from the input.
         """
-        for input in inputs:
-            yield self.extract_one(input)
+        return map(self.extract_one, inputs)