Custom error handling in request strategies

[pitrafa]

Feature Request: A possible hook for custom error handling in request strategies

Is your feature request related to a problem? Please describe. When I want to implement custom logic for handling specific request errors, like timeouts or API rate limits, I find that I need to override the entire _make_request method in my RequestStrategy subclass.

The current implementation in RequestStrategy handles exceptions inside _make_request:

except Exception as error:
    raise UnexpectedError(f"Error when contacting '{endpoint}'") from error

To change this behavior, I have to copy the rest of the _make_request method's logic into my subclass. This works, but it can be a bit cumbersome and means my implementation might get out of sync if the base _make_request method is updated in the future.

Describe the solution you'd like I was thinking that it might be helpful to have a dedicated method for handling these errors, which could be overridden by subclasses.

Perhaps we could introduce a new method, something like _handle_request_error, in the BaseRequestStrategy. The _make_request method would then call this new method in its except block:

# In RequestStrategy._make_request
try:
    # ... existing request code ...
except Exception as error:
    self._handle_request_error(error, endpoint)

The default implementation in BaseRequestStrategy could keep the exact same behavior as today, so nothing would break for existing users:

# In BaseRequestStrategy
def _handle_request_error(self, error: Exception, endpoint: str) -> None:
    raise UnexpectedError(f"Error when contacting '{endpoint}'") from error

This way, if someone wanted to add custom handling for a specific error, they could just override the new _handle_request_error method. For example:

from requests.exceptions import ReadTimeout

class CustomRequestStrategy(RequestStrategy):
    def _handle_request_error(self, error: Exception, endpoint: str) -> None:
        if isinstance(error, ReadTimeout):
            # My custom handling for timeouts
            raise TimeoutError(f"Request to {endpoint} timed out.") from error
        
        # For all other errors, I can just use the default behavior
        super()._handle_request_error(error, endpoint)

Describe alternatives you've considered As mentioned, the main alternative today is to override the whole _make_request method. This is a valid approach, but this suggestion is aimed at making the process a little easier and reducing code duplication for users who only want to tweak the error handling.

Additional context I think a change like this could be useful for handling different kinds of errors that might require special logic, like:

• Request timeouts
• Rate limiting responses
• Specific network issues
• API-specific error formats

This is just a suggestion, and I'm sure you would have a better idea of the best way to implement it. Thank you for considering this feature request