anyio broken resource error with asyncio as backend

[Kinuseka]

Iam not sure if whether this is an anyio or httpx issue, but since this error only occurs during the usage of httpx, then I will place it here.

Though the download is success, but when theres connection error on the way, (which is collected as timeout) the program returns an error but still outputs the correct data anyway.
Here are 2 runnable codes:

Using ASYNCIO:

import httpx
import anyio
import asyncio

async def download(file,link,client):
  full_name = "%s.jpg" % file
  timeout = 0
  while True:
    try:
      async with client.stream('GET',link) as resp:
        async with await anyio.open_file(full_name,"wb") as asf:
          async for chunk in resp.aiter_bytes(1024):
            await asf.write(chunk)
          return True
    except httpx.HTTPError as e:
      timeout += 1 
      if timeout < 6:
        print("%s is retrying %s/5" % (full_name,timeout))
        await asyncio.sleep(2)
        continue
      return False
      print("%s is failure" % full_name)
    

async def main():
  link = "http://kinuseka.github.io/server/background.jpg"
  task = []
  async with httpx.AsyncClient() as client:
    for num in range(1,11):
      task.append(download(num,link,client))
    status = await asyncio.gather(*task)
    if status.count(True) == len(status):
      print("Perfect Request")
    else:
      print("Unfinished Request")
      
    
if __name__ == "__main__":
  print("Running")
  asyncio.run(main())

Using ANYIO(WITH TRIO BACKEND):

import httpx
import anyio
import sys
import asyncio

async def download(file, link, client):
    full_name = "%s.jpg" % file
    timeout = 0 
    print(full_name)
    while True:
        try:
            async with client.stream("GET", link) as resp, await anyio.open_file(
                full_name, "wb"
            ) as asf:
                async for chunk in resp.aiter_bytes(1024):
                    await asf.write(chunk)
                return True
        except httpx.HTTPError as e:
            timeout += 1
            if timeout < 6:
                print("%s is retrying %s/5 (%s)" % (full_name, timeout,e))
                await anyio.sleep(1)
                continue
            return False
            print("%s is failure" % full_name)


async def amain():
    print("Running")
    link = "http://kinuseka.github.io/server/background.jpg"
    perfect_request = True
    async def _do_download(*args):
        if not await download(*args):
            perfect_request = False

    async with httpx.AsyncClient() as client, anyio.create_task_group() as tg:
        for num in range(1, 11):
            tg.start_soon(_do_download, num, link, client)

    if perfect_request:
        print("Perfect Request")
        return 0

    print("Unfinished Request")
    return 1


def main():
  return anyio.run(amain, backend="asyncio")


if __name__ == "__main__":
    sys.exit(main())

Whenever an http error occurs during the async process, the request is given a chance to collect and wait into the except line before eventually retrying the request, the problem with it is that the code spams with a bunch of tracebacks depending how many times http errors has occured
Example:

Future exception was never retrieved
future: <Future finished exception=BrokenResourceError()>
Traceback (most recent call last):
  File "/data/data/com.termux/files/usr/lib/python3.9/asyncio/selector_events.py", line 856, in _read_ready__data_received
    data = self._sock.recv(self.max_size)
ConnectionAbortedError: [Errno 103] Software caused connection abort

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "/data/data/com.termux/files/usr/lib/python3.9/site-packages/httpcore/_backends/anyio.py", line 60, in read
    return await self.stream.receive(n)
  File "/data/data/com.termux/files/usr/lib/python3.9/site-packages/anyio/streams/tls.py", line 152, in receive
    data = await self._call_sslobject_method(self._ssl_object.read, max_bytes)
  File "/data/data/com.termux/files/usr/lib/python3.9/site-packages/anyio/streams/tls.py", line 109, in _call_sslobject_method
    data = await self.transport_stream.receive()
  File "/data/data/com.termux/files/usr/lib/python3.9/site-packages/anyio/_backends/_asyncio.py", line 1095, in receive
    raise self._protocol.exception
anyio.BrokenResourceError
Future exception was never retrieved
future: <Future finished exception=BrokenResourceError()>
Traceback (most recent call last):
  File "/data/data/com.termux/files/usr/lib/python3.9/asyncio/selector_events.py", line 856, in _read_ready__data_received
    data = self._sock.recv(self.max_size)
ConnectionAbortedError: [Errno 103] Software caused connection abort

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "/data/data/com.termux/files/usr/lib/python3.9/site-packages/httpcore/_backends/anyio.py", line 60, in read
    return await self.stream.receive(n)
  File "/data/data/com.termux/files/usr/lib/python3.9/site-packages/anyio/streams/tls.py", line 152, in receive
    data = await self._call_sslobject_method(self._ssl_object.read, max_bytes)
  File "/data/data/com.termux/files/usr/lib/python3.9/site-packages/anyio/streams/tls.py", line 109, in _call_sslobject_method
    data = await self.transport_stream.receive()
  File "/data/data/com.termux/files/usr/lib/python3.9/site-packages/anyio/_backends/_asyncio.py", line 1095, in receive
    raise self._protocol.exception
anyio.BrokenResourceError

HOWEVER! using trio as a backend fixes this issue.

Iam posting this here because I know this is not an expected behavior
I am not sure if this issue is from httpx or anyio but the error occured within the httpx code

[NeyVisQi]

I've had this problem directly with asyncio and httpx before.
In my case, anyio's upgrade fixed the problem.