Leon/docutranslate
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

修复可能的ai无限回复时出现的问题

Browse Source
This commit is contained in:
xunbu
2026-01-06 19:45:55 +08:00
parent 8127ef8267
commit dfa94aac9d
1 changed files with 148 additions and 58 deletions
Download Patch File Download Diff File Expand all files Collapse all files

166
docutranslate/agents/agent.py
View File

@@ -3,6 +3,7 @@
import asyncio
import itertools
import json
import logging
import time
from collections import deque
@@ -21,6 +22,7 @@ from docutranslate.logger import global_logger
from docutranslate.utils.utils import get_httpx_proxies
MAX_REQUESTS_PER_ERROR = 15
MAX_CONTINUE_FETCHES = 2 # 响应被截断时，最多继续获取的次数
ThinkingMode = Literal["enable", "disable", "default"]
@@ -370,15 +372,27 @@ class Agent:
error_result_handler: ErrorResultHandlerType,
retry_count: int,
accumulated_result: str = "",
continue_count: int = 0,
) -> Any:
"""
当 finish_reason 为 length 时，继续获取剩余内容
当 finish_reason 为 length 时，继续获取剩余内容。
注意：很多 API 并不支持这种"继续获取"模式，可能直接返回 stop 或不返回 length。
本方法具有退化机制：如果 API 不支持继续获取，会返回已累计的结果。
最多继续获取 MAX_CONTINUE_FETCHES 次，防止无限循环。
"""
self.logger.info(f"继续获取剩余内容 (已累计 {len(accumulated_result)} 字符)...")
if continue_count >= MAX_CONTINUE_FETCHES:
self.logger.warning(f"已达到最大继续获取次数 ({MAX_CONTINUE_FETCHES})，返回已累计结果 ({len(accumulated_result)} 字符)")
return (
accumulated_result
if result_handler is None
else result_handler(accumulated_result, prompt, self.logger)
)
# 使用空内容继续请求，实际上多数 API 需要用户提供已获取的内容作为上下文
# 这里我们发送一个继续信号，让模型继续输出
continue_prompt = f"{prompt}\n\n[系统提示：之前的响应被截断，请继续输出剩余内容。]"
self.logger.info(f"继续获取剩余内容 (已累计 {len(accumulated_result)} 字符, 第 {continue_count + 1}/{MAX_CONTINUE_FETCHES} 次)...")
# 构造继续请求的提示
# 关键：告知模型我们已经获取了部分内容，请继续完成
continue_prompt = f"{prompt}\n\n[系统提示：请继续完成之前的响应。之前已输出内容为：\n---\n{accumulated_result}\n---\n请从中断处继续输出剩余内容。]"
if pre_send_handler:
system_prompt, continue_prompt = pre_send_handler(system_prompt, continue_prompt)
@@ -399,8 +413,16 @@ class Agent:
response.raise_for_status()
response_data = response.json()
finish_reason = response_data.get("choices", [{}])[0].get("finish_reason", None)
additional_result = response_data["choices"][0]["message"]["content"]
# 安全提取 choices 和 content
choices = response_data.get("choices", [])
if not choices:
self.logger.error(f"API响应中未找到 choices 字段")
raise ValueError("API响应格式错误：缺少 choices 字段")
choice = choices[0]
finish_reason = choice.get("finish_reason", None)
message = choice.get("message", {})
additional_result = message.get("content", "")
input_tokens, cached_tokens, output_tokens, reasoning_tokens = (
extract_token_info(response_data)
@@ -410,7 +432,7 @@ class Agent:
# 累加结果
accumulated_result += additional_result
# 如果仍然是 length，继续获取
# 如果仍然是 length，继续获取（限制最大轮数防止无限循环）
if finish_reason == "length":
return await self._continue_fetch_async(
client=client,
@@ -422,6 +444,7 @@ class Agent:
error_result_handler=error_result_handler,
retry_count=retry_count,
accumulated_result=accumulated_result,
continue_count=continue_count + 1,
)
# 非 length 结束，返回累加结果
@@ -433,8 +456,9 @@ class Agent:
except (httpx.HTTPStatusError, httpx.RequestError, KeyError, IndexError, ValueError) as e:
self.logger.error(f"继续获取内容失败: {repr(e)}")
# 返回已获取的部分结果
# 退化：返回已获取的部分结果，而不是报错
if accumulated_result:
self.logger.warning(f"API不支持继续获取，返回已获取的部分结果 ({len(accumulated_result)} 字符)")
return (
accumulated_result
if result_handler is None
@@ -489,15 +513,21 @@ class Agent:
response_data = response.json()
# 检查 finish_reason
finish_reason = response_data.get("choices", [{}])[0].get("finish_reason", None)
if finish_reason != "stop":
# 非正常结束，可能是 length (长度限制)、tool_calls、content_filter 等
self.logger.warning(
f"finish_reason 为 '{finish_reason}'，非正常结束。prompt: {prompt[:100]}..."
)
choices = response_data.get("choices", [])
if not choices:
self.logger.error(f"API响应中未找到 choices 字段")
raise ValueError("API响应格式错误：缺少 choices 字段")
# 如果是长度限制，尝试继续获取
if finish_reason == "length":
finish_reason = choices[0].get("finish_reason", None)
result = choices[0].get("message", {}).get("content", "")
# 处理不同的 finish_reason
if finish_reason == "stop":
# 正常结束
pass
elif finish_reason == "length":
# 长度限制，尝试继续获取
self.logger.warning(f"响应因长度限制被截断，尝试继续获取...")
return await self._continue_fetch_async(
client=client,
prompt=prompt,
@@ -507,9 +537,25 @@ class Agent:
result_handler=result_handler,
error_result_handler=error_result_handler,
retry_count=retry_count,
accumulated_result=result,
)
result = response_data["choices"][0]["message"]["content"]
elif finish_reason in ("tool_calls", "function_call"):
# 工具调用场景，当前代码可能不支持，直接返回已获取结果
self.logger.warning(f"finish_reason 为 '{finish_reason}'，当前不支持工具调用，返回已获取内容")
return result if result else (
prompt if error_result_handler is None
else error_result_handler(prompt, self.logger)
)
elif finish_reason == "content_filter":
# 内容被过滤
self.logger.error(f"响应内容被过滤")
raise ValueError("内容被过滤")
elif finish_reason is None:
# 某些 API 可能不返回 finish_reason，将其视为正常结束
self.logger.warning(f"API未返回 finish_reason，视为正常结束")
else:
# 其他未知的 finish_reason，记录警告并返回结果
self.logger.warning(f"未知的 finish_reason: '{finish_reason}'，返回已获取内容")
input_tokens, cached_tokens, output_tokens, reasoning_tokens = (
extract_token_info(response_data)
@@ -552,7 +598,7 @@ class Agent:
self.logger.error(f"AI请求连接错误 (async): {repr(e)}")
should_retry = True
is_hard_error = True
except (KeyError, IndexError, ValueError) as e:
except (KeyError, IndexError, ValueError, json.JSONDecodeError) as e:
self.logger.error(f"AI响应格式或值错误 (async), 将尝试重试: {repr(e)}")
should_retry = True
is_hard_error = True
@@ -714,13 +760,26 @@ class Agent:
error_result_handler,
retry_count: int,
accumulated_result: str = "",
continue_count: int = 0,
) -> Any:
"""
当 finish_reason 为 length 时，继续获取剩余内容（同步版本）
当 finish_reason 为 length 时，继续获取剩余内容（同步版本）。
注意：很多 API 并不支持这种"继续获取"模式，可能直接返回 stop 或不返回 length。
本方法具有退化机制：如果 API 不支持继续获取，会返回已累计的结果。
最多继续获取 MAX_CONTINUE_FETCHES 次，防止无限循环。
"""
self.logger.info(f"继续获取剩余内容 (已累计 {len(accumulated_result)} 字符)...")
if continue_count >= MAX_CONTINUE_FETCHES:
self.logger.warning(f"已达到最大继续获取次数 ({MAX_CONTINUE_FETCHES})，返回已累计结果 ({len(accumulated_result)} 字符)")
return (
accumulated_result
if result_handler is None
else result_handler(accumulated_result, prompt, self.logger)
)
continue_prompt = f"{prompt}\n\n[系统提示：之前的响应被截断，请继续输出剩余内容。]"
self.logger.info(f"继续获取剩余内容 (已累计 {len(accumulated_result)} 字符, 第 {continue_count + 1}/{MAX_CONTINUE_FETCHES} 次)...")
# 构造继续请求的提示
continue_prompt = f"{prompt}\n\n[系统提示：请继续完成之前的响应。之前已输出内容为：\n---\n{accumulated_result}\n---\n请从中断处继续输出剩余内容。]"
if pre_send_handler:
system_prompt, continue_prompt = pre_send_handler(system_prompt, continue_prompt)
@@ -740,8 +799,16 @@ class Agent:
response.raise_for_status()
response_data = response.json()
finish_reason = response_data.get("choices", [{}])[0].get("finish_reason", None)
additional_result = response_data["choices"][0]["message"]["content"]
# 安全提取 choices 和 content
choices = response_data.get("choices", [])
if not choices:
self.logger.error(f"API响应中未找到 choices 字段")
raise ValueError("API响应格式错误：缺少 choices 字段")
choice = choices[0]
finish_reason = choice.get("finish_reason", None)
message = choice.get("message", {})
additional_result = message.get("content", "")
input_tokens, cached_tokens, output_tokens, reasoning_tokens = (
extract_token_info(response_data)
@@ -761,6 +828,7 @@ class Agent:
error_result_handler=error_result_handler,
retry_count=retry_count,
accumulated_result=accumulated_result,
continue_count=continue_count + 1,
)
return (
@@ -771,7 +839,9 @@ class Agent:
except (httpx.HTTPStatusError, httpx.RequestError, KeyError, IndexError, ValueError) as e:
self.logger.error(f"继续获取内容失败: {repr(e)}")
# 退化：返回已获取的部分结果，而不是报错
if accumulated_result:
self.logger.warning(f"API不支持继续获取，返回已获取的部分结果 ({len(accumulated_result)} 字符)")
return (
accumulated_result
if result_handler is None
@@ -809,8 +879,6 @@ class Agent:
should_retry = False
is_hard_error = False
current_partial_result = None
input_tokens = 0
output_tokens = 0
try:
response = client.post(
@@ -823,15 +891,21 @@ class Agent:
response_data = response.json()
# 检查 finish_reason
finish_reason = response_data.get("choices", [{}])[0].get("finish_reason", None)
if finish_reason != "stop":
# 非正常结束，可能是 length (长度限制)、tool_calls、content_filter 等
self.logger.warning(
f"finish_reason 为 '{finish_reason}'，非正常结束。prompt: {prompt[:100]}..."
)
choices = response_data.get("choices", [])
if not choices:
self.logger.error(f"API响应中未找到 choices 字段")
raise ValueError("API响应格式错误：缺少 choices 字段")
# 如果是长度限制，尝试继续获取
if finish_reason == "length":
finish_reason = choices[0].get("finish_reason", None)
result = choices[0].get("message", {}).get("content", "")
# 处理不同的 finish_reason
if finish_reason == "stop":
# 正常结束
pass
elif finish_reason == "length":
# 长度限制，尝试继续获取
self.logger.warning(f"响应因长度限制被截断，尝试继续获取...")
return self._continue_fetch(
client=client,
prompt=prompt,
@@ -841,9 +915,25 @@ class Agent:
result_handler=result_handler,
error_result_handler=error_result_handler,
retry_count=retry_count,
accumulated_result=result,
)
result = response_data["choices"][0]["message"]["content"]
elif finish_reason in ("tool_calls", "function_call"):
# 工具调用场景，当前代码可能不支持，直接返回已获取结果
self.logger.warning(f"finish_reason 为 '{finish_reason}'，当前不支持工具调用，返回已获取内容")
return result if result else (
prompt if error_result_handler is None
else error_result_handler(prompt, self.logger)
)
elif finish_reason == "content_filter":
# 内容被过滤
self.logger.error(f"响应内容被过滤")
raise ValueError("内容被过滤")
elif finish_reason is None:
# 某些 API 可能不返回 finish_reason，将其视为正常结束
self.logger.warning(f"API未返回 finish_reason，视为正常结束")
else:
# 其他未知的 finish_reason，记录警告并返回结果
self.logger.warning(f"未知的 finish_reason: '{finish_reason}'，返回已获取内容")
input_tokens, cached_tokens, output_tokens, reasoning_tokens = (
extract_token_info(response_data)
@@ -882,7 +972,7 @@ class Agent:
self.logger.error(f"AI请求连接错误 (sync): {repr(e)}\nprompt:{prompt}")
should_retry = True
is_hard_error = True
except (KeyError, IndexError, ValueError) as e:
except (KeyError, IndexError, ValueError, json.JSONDecodeError) as e:
self.logger.error(f"AI响应格式或值错误 (sync), 将尝试重试: {repr(e)}")
should_retry = True
is_hard_error = True