Make exceptions raise from the previously caught exception

[Daniël van Noord]

Currently opnieuw will only raise the last caught exception. This creates two issues:

1. We lose any exceptions that were previously raised. This is especially problematic when the first caught exception is different from the second caught exception.
2. It is hard to tell whether opnieuw actually retried in a Sentry stack trace as the retrying isn't explicitly in the trace.

from opnieuw import retry

counter = 0


@retry(
    retry_on_exceptions=(TypeError, ValueError),
    max_calls_total=3,
    retry_window_after_first_call_in_seconds=1,
)
def raise_depending_on_counter() -> None:
    global counter
    counter += 1
    if counter == 1:
        raise TypeError
    elif counter == 2:
        raise ValueError
    else:
        raise IndexError


raise_depending_on_counter()

This currently raises:

❯ python test.py
Traceback (most recent call last):
  File "/home/daniel/channable/requestmachine/test.py", line 22, in <module>
    raise_depending_on_counter()
  File "/nix/store/mzflaklx9497sffvm5k28cf6m0zdi0yd-python3-3.11.4-env/lib/python3.11/site-packages/opnieuw/retries.py", line 240, in wrapper
    return f(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^
  File "/home/daniel/channable/requestmachine/test.py", line 19, in raise_depending_on_counter
    raise IndexError
IndexError

Ideally it would be something like:

❯ python test.py
Traceback (most recent call last):
  File "/nix/store/mzflaklx9497sffvm5k28cf6m0zdi0yd-python3-3.11.4-env/lib/python3.11/site-packages/opnieuw/retries.py", line 241, in wrapper
    return f(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^
  File "/home/daniel/channable/requestmachine/test.py", line 15, in raise_depending_on_counter
    raise TypeError
TypeError

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

Traceback (most recent call last):
  File "/nix/store/mzflaklx9497sffvm5k28cf6m0zdi0yd-python3-3.11.4-env/lib/python3.11/site-packages/opnieuw/retries.py", line 241, in wrapper
    return f(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^
  File "/home/daniel/channable/requestmachine/test.py", line 17, in raise_depending_on_counter
    raise ValueError
ValueError

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

Traceback (most recent call last):
  File "/home/daniel/channable/requestmachine/test.py", line 22, in <module>
    raise_depending_on_counter()
  File "/nix/store/mzflaklx9497sffvm5k28cf6m0zdi0yd-python3-3.11.4-env/lib/python3.11/site-packages/opnieuw/retries.py", line 265, in wrapper
    raise last_exception
  File "/nix/store/mzflaklx9497sffvm5k28cf6m0zdi0yd-python3-3.11.4-env/lib/python3.11/site-packages/opnieuw/retries.py", line 241, in wrapper
    return f(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^
  File "/home/daniel/channable/requestmachine/test.py", line 19, in raise_depending_on_counter
    raise IndexError
IndexError

A draft PR is here: #21.

[Jochem Bijlard]

This sounds like a very useful improvement to the logs in Sentry. This should make it much easier to see what has actually happened before we failed. Using exception chaining via the __cause__ seems like it matches the intended semantics as defined in the PEP3134: Explicit Exception Chaining.