Best way to handle Python errors in a function

Question

We got a serverless function and split the functionality in multiple smaller functions. One function transforms arrays of a dictionary in strings. This is because Azure Datafactory can't handle arrays with the build in copy activity. This is the start code:

def refactor_arrays(data, firma):
        for index, item in enumerate(data):
            data[index]['versionedRepresentations']['components']['1'] = str(
                item['versionedRepresentations']['components']['1'])

            data[index]['versionedRepresentations']['fixVersions']['1'] = str(
                item['versionedRepresentations']['fixVersions']['1'])

            data[index]['versionedRepresentations']['labels']['1'] = str(
                item['versionedRepresentations']['labels']['1'])

            data[index]['versionedRepresentations']['customfield_10005']['2'] = str(
                item['versionedRepresentations']['customfield_10005']['2'])

    return data

This is how I as a bginner would handle errors.

    def refactor_arrays(data, firma):
        for index, item in enumerate(data):
            try:
                data[index]['versionedRepresentations']['components']['1'] = str(
                    item['versionedRepresentations']['components']['1'])
            except KeyError:
                logging.info(
                    'Key does not exist. Most likely another API endpoint gets called.')
            except:
                logging.error(
                    'Something went wrong when trying to convert an array to a string.')

            try:
                data[index]['versionedRepresentations']['fixVersions']['1'] = str(
                    item['versionedRepresentations']['fixVersions']['1'])
            except KeyError:
                logging.info(
                    'Key does not exist. Most likely another API endpoint gets called.')
            except:
                logging.error(
                    'Something went wrong when trying to convert an array to a string.')

            try:
                data[index]['versionedRepresentations']['labels']['1'] = str(
                    item['versionedRepresentations']['labels']['1'])
            except KeyError:
                logging.info(
                    'Key does not exist. Most likely another API endpoint gets called.')
            except:
                logging.error(
                    'Something went wrong when trying to convert an array to a string.')

            try:
                data[index]['versionedRepresentations']['customfield_10005']['2'] = str(
                    item['versionedRepresentations']['customfield_10005']['2'])
            except KeyError:
                logging.info(
                    'Key does not exist. Most likely another API endpoint gets called.')
            except:
                logging.error(
                    'Something went wrong when trying to convert an array to a string.')

    return data

I suggest to put a check for keys in dictionary before accessing them, that way you know the problem and log it. Then put a common try catch block for whole function, no need to put try catch for every statement — Anurag
– Anurag, Commented Feb 13, 2020 at 7:04
Never have have an empty except block as it will also detect system interrupts and keyboard interrupts as they inherit from the base class. — kinshukdua
– kinshukdua, Commented Feb 13, 2020 at 7:13

Flux · Accepted Answer · 2020-02-14 07:03:17Z

You could write a function (e.g. convert_to_string) whose job is (1) to change the values to str and (2) log errors if necessary. Then you can use convert_to_string in refactor_arrays, avoiding code repetition:

def convert_to_string(data, index, a, b, c):
    try:
        data[index][a][b][c] = str(data[index][a][b][c])
    except KeyError:
        logging.info('Key does not exist. Most likely another API endpoint gets called.')
    except:
        logging.error('Something went wrong when trying to convert an array to a string.')


def refactor_arrays(data, firma):
    for index, item in enumerate(data):
        convert_to_string(data, index, 'versionedRepresentations', 'components', '1')
        convert_to_string(data, index, 'fixVersions', 'labels', '1')
        convert_to_string(data, index, 'versionedRepresentations', 'labels', '1')
        convert_to_string(data, index, 'versionedRepresentations', 'customfield_10005', '2')
    return data

Collectives™ on Stack Overflow

Best way to handle Python errors in a function

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related