feat(workflow): domain model for workflow node execution (#19430)

Signed-off-by: -LAN- <laipz8200@outlook.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
This commit is contained in:
-LAN-
2025-05-17 00:56:16 +08:00
committed by GitHub
parent aeceb200ec
commit 4977bb21ec
31 changed files with 1108 additions and 483 deletions

View File

@@ -1,3 +1,4 @@
from collections.abc import Mapping
from datetime import datetime
from enum import StrEnum
from typing import Any, Optional, Union
@@ -30,8 +31,8 @@ class LangSmithMultiModel(BaseModel):
class LangSmithRunModel(LangSmithTokenUsage, LangSmithMultiModel):
name: Optional[str] = Field(..., description="Name of the run")
inputs: Optional[Union[str, dict[str, Any], list, None]] = Field(None, description="Inputs of the run")
outputs: Optional[Union[str, dict[str, Any], list, None]] = Field(None, description="Outputs of the run")
inputs: Optional[Union[str, Mapping[str, Any], list, None]] = Field(None, description="Inputs of the run")
outputs: Optional[Union[str, Mapping[str, Any], list, None]] = Field(None, description="Outputs of the run")
run_type: LangSmithRunType = Field(..., description="Type of the run")
start_time: Optional[datetime | str] = Field(None, description="Start time of the run")
end_time: Optional[datetime | str] = Field(None, description="End time of the run")

View File

@@ -1,4 +1,3 @@
import json
import logging
import os
import uuid
@@ -7,7 +6,7 @@ from typing import Optional, cast
from langsmith import Client
from langsmith.schemas import RunBase
from sqlalchemy.orm import sessionmaker
from sqlalchemy.orm import Session, sessionmaker
from core.ops.base_trace_instance import BaseTraceInstance
from core.ops.entities.config_entity import LangSmithConfig
@@ -29,8 +28,10 @@ from core.ops.langsmith_trace.entities.langsmith_trace_entity import (
)
from core.ops.utils import filter_none_values, generate_dotted_order
from core.repositories import SQLAlchemyWorkflowNodeExecutionRepository
from core.workflow.entities.node_entities import NodeRunMetadataKey
from core.workflow.nodes.enums import NodeType
from extensions.ext_database import db
from models.model import EndUser, MessageFile
from models import Account, App, EndUser, MessageFile, WorkflowNodeExecutionTriggeredFrom
logger = logging.getLogger(__name__)
@@ -137,8 +138,29 @@ class LangSmithDataTrace(BaseTraceInstance):
# through workflow_run_id get all_nodes_execution using repository
session_factory = sessionmaker(bind=db.engine)
# Find the app's creator account
with Session(db.engine, expire_on_commit=False) as session:
# Get the app to find its creator
app_id = trace_info.metadata.get("app_id")
if not app_id:
raise ValueError("No app_id found in trace_info metadata")
app = session.query(App).filter(App.id == app_id).first()
if not app:
raise ValueError(f"App with id {app_id} not found")
if not app.created_by:
raise ValueError(f"App with id {app_id} has no creator (created_by is None)")
service_account = session.query(Account).filter(Account.id == app.created_by).first()
if not service_account:
raise ValueError(f"Creator account with id {app.created_by} not found for app {app_id}")
workflow_node_execution_repository = SQLAlchemyWorkflowNodeExecutionRepository(
session_factory=session_factory, tenant_id=trace_info.tenant_id, app_id=trace_info.metadata.get("app_id")
session_factory=session_factory,
user=service_account,
app_id=trace_info.metadata.get("app_id"),
triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
)
# Get all executions for this workflow run
@@ -148,27 +170,23 @@ class LangSmithDataTrace(BaseTraceInstance):
for node_execution in workflow_node_executions:
node_execution_id = node_execution.id
tenant_id = node_execution.tenant_id
app_id = node_execution.app_id
tenant_id = trace_info.tenant_id # Use from trace_info instead
app_id = trace_info.metadata.get("app_id") # Use from trace_info instead
node_name = node_execution.title
node_type = node_execution.node_type
status = node_execution.status
if node_type == "llm":
inputs = (
json.loads(node_execution.process_data).get("prompts", {}) if node_execution.process_data else {}
)
if node_type == NodeType.LLM:
inputs = node_execution.process_data.get("prompts", {}) if node_execution.process_data else {}
else:
inputs = json.loads(node_execution.inputs) if node_execution.inputs else {}
outputs = json.loads(node_execution.outputs) if node_execution.outputs else {}
inputs = node_execution.inputs if node_execution.inputs else {}
outputs = node_execution.outputs if node_execution.outputs else {}
created_at = node_execution.created_at or datetime.now()
elapsed_time = node_execution.elapsed_time
finished_at = created_at + timedelta(seconds=elapsed_time)
execution_metadata = (
json.loads(node_execution.execution_metadata) if node_execution.execution_metadata else {}
)
node_total_tokens = execution_metadata.get("total_tokens", 0)
metadata = execution_metadata.copy()
execution_metadata = node_execution.metadata if node_execution.metadata else {}
node_total_tokens = execution_metadata.get(NodeRunMetadataKey.TOTAL_TOKENS) or 0
metadata = {str(key): value for key, value in execution_metadata.items()}
metadata.update(
{
"workflow_run_id": trace_info.workflow_run_id,
@@ -181,7 +199,7 @@ class LangSmithDataTrace(BaseTraceInstance):
}
)
process_data = json.loads(node_execution.process_data) if node_execution.process_data else {}
process_data = node_execution.process_data if node_execution.process_data else {}
if process_data and process_data.get("model_mode") == "chat":
run_type = LangSmithRunType.llm
@@ -191,7 +209,7 @@ class LangSmithDataTrace(BaseTraceInstance):
"ls_model_name": process_data.get("model_name", ""),
}
)
elif node_type == "knowledge-retrieval":
elif node_type == NodeType.KNOWLEDGE_RETRIEVAL:
run_type = LangSmithRunType.retriever
else:
run_type = LangSmithRunType.tool