aiorpcx: fix bug in timeout_after monkeypatch

fixes https://github.com/spesmilo/electrum/issues/7952

see in particular https://github.com/spesmilo/electrum/issues/7952#issuecomment-1227225602
> So the issue is with the aiorpcx monkey patch in util.py, as it
> relies on side-effecting the asyncio.Task, and it patches Task.cancel.
> However, aiohttp also uses Task.cancel for its own timeouting of the
> http request, with the same Task object, and this confuses timeout_after.
> Ultimately FxThread.run exits.

related https://github.com/kyuupichan/aiorpcX/pull/47

---

note: I am not content at all with this monkey-patching approach,
but at the same time I don't see how to properly fix this handling all
edge-cases in aiorpcx.

python 3.11 is finally adding an implementation of TaskGroup [0] and
an async timeout context manager [1] in the asyncio module of the stdlib.
Looking at the implementations, they look unfeasible to backport:
much of the implementation of asyncio.Task had to be changed for them
to work, and TaskGroup in particular relies on the new ExceptionGroups.
Some of these edge cases we are battling with aiorpcx.curio look
difficult to fix without those stdlib changes...
Anyway, when we bump the min python to 3.11, I look forward to switching
to that code instead, and ripping this stuff out.

[0]: https://docs.python.org/3.11/library/asyncio-task.html#task-groups
[1]: https://docs.python.org/3.11/library/asyncio-task.html#asyncio.timeout

electrum/util.py

@@ -1324,6 +1324,7 @@ class OldTaskGroup(aiorpcx.TaskGroup):
         await group.spawn(task1())
         await group.spawn(task2())
     ```
+    # TODO see if we can migrate to asyncio.timeout, introduced in python 3.11, and use stdlib instead of aiorpcx.curio...
     """
     async def join(self):
         if self._wait is all:
@@ -1360,6 +1361,7 @@ class OldTaskGroup(aiorpcx.TaskGroup):
 # see https://github.com/kyuupichan/aiorpcX/issues/44
 # see https://github.com/aio-libs/async-timeout/issues/229
 # see https://bugs.python.org/issue42130 and https://bugs.python.org/issue45098
+# TODO see if we can migrate to asyncio.timeout, introduced in python 3.11, and use stdlib instead of aiorpcx.curio...
 def _aiorpcx_monkeypatched_set_new_deadline(task, deadline):
     def timeout_task():
         task._orig_cancel()
@@ -1373,7 +1375,26 @@ def _aiorpcx_monkeypatched_set_new_deadline(task, deadline):
         task.cancel = mycancel
     task._deadline_handle = task._loop.call_at(deadline, timeout_task)
 
+
+def _aiorpcx_monkeypatched_set_task_deadline(task, deadline):
+    ret = _aiorpcx_orig_set_task_deadline(task, deadline)
+    task._externally_cancelled = None
+    return ret
+
+
+def _aiorpcx_monkeypatched_unset_task_deadline(task):
+    if hasattr(task, "_orig_cancel"):
+        task.cancel = task._orig_cancel
+        del task._orig_cancel
+    return _aiorpcx_orig_unset_task_deadline(task)
+
+
+_aiorpcx_orig_set_task_deadline    = aiorpcx.curio._set_task_deadline
+_aiorpcx_orig_unset_task_deadline  = aiorpcx.curio._unset_task_deadline
+
 aiorpcx.curio._set_new_deadline    = _aiorpcx_monkeypatched_set_new_deadline
+aiorpcx.curio._set_task_deadline   = _aiorpcx_monkeypatched_set_task_deadline
+aiorpcx.curio._unset_task_deadline = _aiorpcx_monkeypatched_unset_task_deadline
 
 
 class NetworkJobOnDefaultServer(Logger, ABC):