I was debugging an additional pipeline function with the steam backend of python-social-auth 0.1.17, and was seeing an abort of the login pipeline whenever my function ran (which failed to log in the user). I saw no errors in my logs, but eventually discovered that my function was raising an exception. It appeared that the exception was being taken as an HTTP response. This makes sense under the literal interpretation of "non-dictionary responses will be taken as HTTP responses," but is counterintuitive and promotes wrapping all your functions in try-except block to get logging. Is there a more maintainable way of seeing when exceptions are occurring, and perhaps differentiating exceptions from normal function returns?
I was debugging an additional pipeline function with the steam backend of python-social-auth 0.1.17, and was seeing an abort of the login pipeline whenever my function ran (which failed to log in the user). I saw no errors in my logs, but eventually discovered that my function was raising an exception. It appeared that the exception was being taken as an HTTP response. This makes sense under the literal interpretation of "non-dictionary responses will be taken as HTTP responses," but is counterintuitive and promotes wrapping all your functions in try-except block to get logging. Is there a more maintainable way of seeing when exceptions are occurring, and perhaps differentiating exceptions from normal function returns?