如何修复使用asyncio+aiohttp并行请求时的Session Closed错误？

我尝试编写一个Python脚本，用于发起批量并行HTTP请求，以下是我的代码实现：

async def deleteLinkRecords(req_headers):
  async with aiohttp.ClientSession(trust_env=True) as session:
   idm_url = "https://example.com"

   tasks = []
   for id in linkIdArr:
     del_url = idm_url + id
     tasks.append(make_request(session, del_url, req_headers))

  results = await asyncio.gather(*tasks)
  print(results)

async def make_request(session, url, req_headers):
    async with session.delete(url = url, headers = req_headers) as resp:
     return await resp.json()

my_headers = {
   "Authorization": "Bearer XXX"
}

linkIdArr = ["abc", "def"....]

asyncio.run(deleteLinkRecords(req_headers = my_headers))

运行该脚本时，我遇到了如下错误：

Traceback (most recent call last):
  File "C:\Users\mycode.py", line 48, in <module>
    asyncio.run(deleteLinkRecords(req_headers = my_headers))
  File "C:\Users\AppData\Local\Programs\Python\Python39\lib\asyncio\runners.py", line 44, in run
    return loop.run_until_complete(main)
  File "C:\Users\AppData\Local\Programs\Python\Python39\lib\asyncio\base_events.py", line 642, in run_until_complete
    return future.result()
  File "C:\Users\mycode.py", line 20, in deleteLinkRecords
    results = await asyncio.gather(*tasks)
  File "C:\Users\mycode.py", line 28, in make_requests
    async with session.delete(url = url, headers = req_headers) as resp:
  File "C:\Users\AppData\Local\Programs\Python\Python39\lib\site-packages\aiohttp\client.py", line 1197, in __aenter__
    self._resp = await self._coro
  File "C:\Users\AppData\Local\Programs\Python\Python39\lib\site-packages\aiohttp\client.py", line 428, in _request
    raise RuntimeError("Session is closed")
RuntimeError: Session is closed

核心问题 #

代码缩进错误导致await asyncio.gather(*tasks)跑到了async with aiohttp.ClientSession的上下文之外。当上下文管理器退出时，session会被立即关闭，此时再执行任务里的请求就会触发"Session is closed"错误。

修正后的完整代码 #

import asyncio
import aiohttp

async def deleteLinkRecords(req_headers):
    async with aiohttp.ClientSession(trust_env=True) as session:
        idm_url = "https://example.com"
        tasks = []
        for link_id in linkIdArr:  # 避免使用内置关键字id作为变量名
            del_url = idm_url + link_id
            tasks.append(make_request(session, del_url, req_headers))
        
        # 将gather调用放在async with代码块内部，确保session在请求执行期间保持打开
        results = await asyncio.gather(*tasks)
        print(results)

async def make_request(session, url, req_headers):
    async with session.delete(url=url, headers=req_headers) as resp:
        return await resp.json()

my_headers = {
    "Authorization": "Bearer XXX"
}

linkIdArr = ["abc", "def"]  # 补充完整的ID列表

asyncio.run(deleteLinkRecords(req_headers=my_headers))

额外优化点 #

1. 避免使用Python内置关键字id作为循环变量名，改用link_id避免潜在问题。
2. 补充了缺失的asyncio和aiohttp导入语句，确保脚本能正常运行。
3. 统一了代码缩进风格，提升可读性。

内容的提问来源于stack exchange，提问作者RKodakandla