mirror of https://github.com/kortix-ai/suna.git
exec on stream as background task
This commit is contained in:
marko-kraemer
parent
d94f87b0f4
commit
6d6dda1284
|
|
@ -9,7 +9,6 @@ This module handles processing of LLM responses including:
|
|||
"""
|
||||
|
||||
import json
|
||||
import logging
|
||||
import asyncio
|
||||
import re
|
||||
import uuid
|
||||
|
|
@ -28,7 +27,8 @@ ToolExecutionStrategy = Literal["sequential", "parallel"]
|
|||
|
||||
@dataclass
|
||||
class ProcessorConfig:
|
||||
"""Configuration for response processing and tool execution.
|
||||
"""
|
||||
Configuration for response processing and tool execution.
|
||||
|
||||
This class controls how the LLM's responses are processed, including how tool calls
|
||||
are detected, executed, and their results handled.
|
||||
|
|
@ -41,11 +41,10 @@ class ProcessorConfig:
|
|||
tool_execution_strategy: How to execute multiple tools ("sequential" or "parallel")
|
||||
xml_adding_strategy: How to add XML tool results to the conversation
|
||||
"""
|
||||
# Tool detection
|
||||
|
||||
xml_tool_calling: bool = True
|
||||
native_tool_calling: bool = False
|
||||
|
||||
# Tool execution
|
||||
execute_tools: bool = True
|
||||
execute_on_stream: bool = False
|
||||
tool_execution_strategy: ToolExecutionStrategy = "sequential"
|
||||
|
|
@ -98,6 +97,9 @@ class ResponseProcessor:
|
|||
# For tracking tool results during streaming to add later
|
||||
tool_results_buffer = []
|
||||
|
||||
# For tracking pending tool executions
|
||||
pending_tool_executions = []
|
||||
|
||||
logger.info(f"Starting to process streaming response for thread {thread_id}")
|
||||
logger.info(f"Config: XML={config.xml_tool_calling}, Native={config.native_tool_calling}, "
|
||||
f"Execute on stream={config.execute_on_stream}, Execution strategy={config.tool_execution_strategy}")
|
||||
|
|
@ -131,19 +133,18 @@ class ResponseProcessor:
|
|||
# Parse and extract the tool call
|
||||
tool_call = self._parse_xml_tool_call(xml_chunk)
|
||||
if tool_call:
|
||||
# Execute tool if needed, but store results for later
|
||||
# Execute tool if needed, but in background
|
||||
if config.execute_tools and config.execute_on_stream:
|
||||
result = await self._execute_tool(tool_call)
|
||||
# Start tool execution as a background task
|
||||
execution_task = asyncio.create_task(self._execute_tool(tool_call))
|
||||
|
||||
# Store result to add after assistant message
|
||||
tool_results_buffer.append((tool_call, result))
|
||||
# Store the task for later retrieval
|
||||
pending_tool_executions.append({
|
||||
"task": execution_task,
|
||||
"tool_call": tool_call
|
||||
})
|
||||
|
||||
# Yield tool execution result for client display
|
||||
yield {
|
||||
"type": "tool_result",
|
||||
"name": tool_call["name"],
|
||||
"result": str(result)
|
||||
}
|
||||
# Immediately continue processing more chunks
|
||||
|
||||
# Process native tool calls
|
||||
if config.native_tool_calling and delta and hasattr(delta, 'tool_calls') and delta.tool_calls:
|
||||
|
|
@ -190,17 +191,73 @@ class ResponseProcessor:
|
|||
"arguments": json.loads(current_tool['function']['arguments']),
|
||||
"id": current_tool['id']
|
||||
}
|
||||
result = await self._execute_tool(tool_call_data)
|
||||
|
||||
# Store result to add after assistant message
|
||||
tool_results_buffer.append((tool_call_data, result))
|
||||
# Start tool execution as a background task
|
||||
execution_task = asyncio.create_task(self._execute_tool(tool_call_data))
|
||||
|
||||
# Yield tool execution result for client display
|
||||
yield {
|
||||
"type": "tool_result",
|
||||
"name": tool_call_data["name"],
|
||||
"result": str(result)
|
||||
}
|
||||
# Store the task for later retrieval
|
||||
pending_tool_executions.append({
|
||||
"task": execution_task,
|
||||
"tool_call": tool_call_data
|
||||
})
|
||||
|
||||
# Immediately continue processing more chunks
|
||||
|
||||
# Check for completed tool executions
|
||||
completed_executions = []
|
||||
for i, execution in enumerate(pending_tool_executions):
|
||||
if execution["task"].done():
|
||||
try:
|
||||
# Get the result
|
||||
result = execution["task"].result()
|
||||
tool_call = execution["tool_call"]
|
||||
|
||||
# Store result for later database updates
|
||||
tool_results_buffer.append((tool_call, result))
|
||||
|
||||
# Yield tool execution result for client display
|
||||
yield {
|
||||
"type": "tool_result",
|
||||
"name": tool_call["name"],
|
||||
"result": str(result)
|
||||
}
|
||||
|
||||
# Mark for removal
|
||||
completed_executions.append(i)
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error getting tool execution result: {str(e)}")
|
||||
|
||||
# Remove completed executions from pending list (in reverse to maintain indices)
|
||||
for i in sorted(completed_executions, reverse=True):
|
||||
pending_tool_executions.pop(i)
|
||||
|
||||
# After streaming completes, wait for any remaining tool executions
|
||||
if pending_tool_executions:
|
||||
logger.info(f"Waiting for {len(pending_tool_executions)} pending tool executions to complete")
|
||||
|
||||
# Wait for all pending tasks to complete
|
||||
pending_tasks = [execution["task"] for execution in pending_tool_executions]
|
||||
done, _ = await asyncio.wait(pending_tasks)
|
||||
|
||||
# Process results
|
||||
for execution in pending_tool_executions:
|
||||
try:
|
||||
if execution["task"].done():
|
||||
result = execution["task"].result()
|
||||
tool_call = execution["tool_call"]
|
||||
|
||||
# Store result for later
|
||||
tool_results_buffer.append((tool_call, result))
|
||||
|
||||
# Yield tool execution result
|
||||
yield {
|
||||
"type": "tool_result",
|
||||
"name": tool_call["name"],
|
||||
"result": str(result)
|
||||
}
|
||||
except Exception as e:
|
||||
logger.error(f"Error processing remaining tool execution: {str(e)}")
|
||||
|
||||
# After streaming completes, process any remaining content and tool calls
|
||||
if accumulated_content:
|
||||
|
|
|
|||
|
|
@ -10,19 +10,14 @@ This module provides comprehensive conversation management, including:
|
|||
"""
|
||||
|
||||
import json
|
||||
import logging
|
||||
import asyncio
|
||||
import uuid
|
||||
import re
|
||||
from typing import List, Dict, Any, Optional, Type, Union, AsyncGenerator, Tuple, Callable, Literal
|
||||
from services.llm import make_llm_api_call
|
||||
from agentpress.tool import Tool, ToolResult
|
||||
from agentpress.tool_registry import ToolRegistry
|
||||
from agentpress.response_processor import (
|
||||
ResponseProcessor,
|
||||
ProcessorConfig,
|
||||
XmlAddingStrategy,
|
||||
ToolExecutionStrategy
|
||||
ProcessorConfig
|
||||
)
|
||||
from services.supabase import DBConnection
|
||||
from utils.logger import logger
|
||||
|
|
@ -318,6 +313,15 @@ class ThreadManager:
|
|||
logger.debug(f"Processor config: XML={processor_config.xml_tool_calling}, Native={processor_config.native_tool_calling}, "
|
||||
f"Execute tools={processor_config.execute_tools}, Strategy={processor_config.tool_execution_strategy}")
|
||||
|
||||
# Check if native_tool_calling is enabled and throw an error if it is
|
||||
if processor_config.native_tool_calling:
|
||||
error_message = "Native tool calling is not supported in this version"
|
||||
logger.error(error_message)
|
||||
return {
|
||||
"status": "error",
|
||||
"message": error_message
|
||||
}
|
||||
|
||||
# 4. Prepare tools for LLM call
|
||||
openapi_tool_schemas = None
|
||||
if processor_config.native_tool_calling:
|
||||
|
|
|
|||
Loading…
Reference in New Issue