Merge branch 'main' into feat/trigger

api/core/workflow/nodes/agent/agent_node.py

@@ -153,7 +153,7 @@ class AgentNode(BaseNode):
                 messages=message_stream,
                 tool_info={
                     "icon": self.agent_strategy_icon,
-                    "agent_strategy": cast(AgentNodeData, self._node_data).agent_strategy_name,
+                    "agent_strategy": self._node_data.agent_strategy_name,
                 },
                 parameters_for_log=parameters_for_log,
                 user_id=self.user_id,
@@ -394,8 +394,7 @@ class AgentNode(BaseNode):
             current_plugin = next(
                 plugin
                 for plugin in plugins
-                if f"{plugin.plugin_id}/{plugin.name}"
-                == cast(AgentNodeData, self._node_data).agent_strategy_provider_name
+                if f"{plugin.plugin_id}/{plugin.name}" == self._node_data.agent_strategy_provider_name
             )
             icon = current_plugin.declaration.icon
         except StopIteration:

api/core/workflow/nodes/answer/answer_stream_processor.py

@@ -149,9 +149,6 @@ class AnswerStreamProcessor(StreamProcessor):
             return []
 
         stream_output_value_selector = event.from_variable_selector
-        if not stream_output_value_selector:
-            return []
-
         stream_out_answer_node_ids = []
         for answer_node_id, route_position in self.route_position.items():
             if answer_node_id not in self.rest_node_ids:

api/core/workflow/nodes/document_extractor/node.py

@@ -302,12 +302,12 @@ def _extract_text_from_yaml(file_content: bytes) -> str:
             encoding = "utf-8"
 
         yaml_data = yaml.safe_load_all(file_content.decode(encoding, errors="ignore"))
-        return cast(str, yaml.dump_all(yaml_data, allow_unicode=True, sort_keys=False))
+        return yaml.dump_all(yaml_data, allow_unicode=True, sort_keys=False)
     except (UnicodeDecodeError, LookupError, yaml.YAMLError) as e:
         # If decoding fails, try with utf-8 as last resort
         try:
             yaml_data = yaml.safe_load_all(file_content.decode("utf-8", errors="ignore"))
-            return cast(str, yaml.dump_all(yaml_data, allow_unicode=True, sort_keys=False))
+            return yaml.dump_all(yaml_data, allow_unicode=True, sort_keys=False)
         except (UnicodeDecodeError, yaml.YAMLError):
             raise TextExtractionError(f"Failed to decode or parse YAML file: {e}") from e
 
@@ -515,14 +515,14 @@ def _extract_text_from_excel(file_content: bytes) -> str:
                 df.dropna(how="all", inplace=True)
 
                 # Combine multi-line text in each cell into a single line
-                df = df.applymap(lambda x: " ".join(str(x).splitlines()) if isinstance(x, str) else x)  # type: ignore
+                df = df.map(lambda x: " ".join(str(x).splitlines()) if isinstance(x, str) else x)
 
                 # Combine multi-line text in column names into a single line
                 df.columns = pd.Index([" ".join(str(col).splitlines()) for col in df.columns])
 
                 # Manually construct the Markdown table
                 markdown_table += _construct_markdown_table(df) + "\n\n"
-            except Exception as e:
+            except Exception:
                 continue
         return markdown_table
     except Exception as e:

api/core/workflow/nodes/knowledge_retrieval/knowledge_retrieval_node.py

@@ -78,7 +78,7 @@ default_retrieval_model = {
     "search_method": RetrievalMethod.SEMANTIC_SEARCH.value,
     "reranking_enable": False,
     "reranking_model": {"reranking_provider_name": "", "reranking_model_name": ""},
-    "top_k": 2,
+    "top_k": 4,
     "score_threshold_enabled": False,
 }
 

api/core/workflow/nodes/loop/loop_node.py

@@ -524,7 +524,10 @@ class LoopNode(BaseNode):
     @staticmethod
     def _get_segment_for_constant(var_type: SegmentType, original_value: Any) -> Segment:
         """Get the appropriate segment type for a constant value."""
-        if not var_type.is_array_type() or var_type == SegmentType.BOOLEAN:
+        # TODO: Refactor for maintainability:
+        # 1. Ensure type handling logic stays synchronized with _VALID_VAR_TYPE (entities.py)
+        # 2. Consider moving this method to LoopVariableData class for better encapsulation
+        if not var_type.is_array_type() or var_type == SegmentType.ARRAY_BOOLEAN:
             value = original_value
         elif var_type in [
             SegmentType.ARRAY_NUMBER,

api/core/workflow/nodes/parameter_extractor/parameter_extractor_node.py

@@ -139,7 +139,7 @@ class ParameterExtractorNode(BaseNode):
         """
         Run the node.
         """
-        node_data = cast(ParameterExtractorNodeData, self._node_data)
+        node_data = self._node_data
         variable = self.graph_runtime_state.variable_pool.get(node_data.query)
         query = variable.text if variable else ""
 

api/core/workflow/nodes/question_classifier/question_classifier_node.py

@@ -1,6 +1,6 @@
 import json
 from collections.abc import Mapping, Sequence
-from typing import TYPE_CHECKING, Any, Optional, cast
+from typing import TYPE_CHECKING, Any, Optional
 
 from core.app.entities.app_invoke_entities import ModelConfigWithCredentialsEntity
 from core.memory.token_buffer_memory import TokenBufferMemory
@@ -109,7 +109,7 @@ class QuestionClassifierNode(BaseNode):
         return "1"
 
     def _run(self):
-        node_data = cast(QuestionClassifierNodeData, self._node_data)
+        node_data = self._node_data
         variable_pool = self.graph_runtime_state.variable_pool
 
         # extract variables

api/core/workflow/nodes/tool/tool_node.py

@@ -1,5 +1,5 @@
 from collections.abc import Generator, Mapping, Sequence
-from typing import Any, Optional, cast
+from typing import Any, Optional
 
 from sqlalchemy import select
 from sqlalchemy.orm import Session
@@ -57,7 +57,7 @@ class ToolNode(BaseNode):
         Run the tool node
         """
 
-        node_data = cast(ToolNodeData, self._node_data)
+        node_data = self._node_data
 
         # fetch tool icon
         tool_info = {