Catch and print full Python exception traceback without halting/exiting the program

Handling errors without blocking the party

To seize the traceback for late-night error analysis, use traceback.format_exc(). It provides a full exception traceback as a string:

import traceback

try:
    oh_no_an_error()
except Exception as e:
    error_traceback = traceback.format_exc()
    # Save for late-night reading or pass it forward


For a more tuned approach when dealing with exceptions, sys.exc_info() provides a tuple containing exception type, value, and traceback detail. Exceptionally useful for custom handling of trivial to catastrophic exceptions.

Advanced exception handling

Use logging for hardcore error tracking

Instead of letting the exceptions scream in the console, make them whisper silently into your logs using logging with logger.exception(). It will broadcast your secrets, but includes the traceback too:

import logging
log = logging.getLogger(__name__)

try:
    code_that_might_explode()
except Exception as error:
    log.exception('Boom! Error Occurred')  # Logging controversies along with traceback


Tidy up with sys.exc_info()

Clear sys.exc_info() references after every rollout to avoid circular references. Even with automatic garbage collection in Python 2.2 and later, it's good practice to maximize your tidiness score:

try:
    # code that could probably stir up a storm
except Exception as e:
    exc_type, exc_value, exc_traceback = sys.exc_info()
    # Handle your storms here
finally:
    del exc_traceback  # Clean up after the party


Get the traceback of your execution journey

To log the state of stack without exceptions use traceback.print_stack(). It's like taking a selfie of your code at its current state:

import traceback

def function_in_the_wild():
    # Curious about who got you here?
    traceback.print_stack()

function_in_the_wild()


Mastering the art of tracebacks

Python 3's secret weapon err.__traceback__

Python 3 unleashes __traceback__ to easily grab the traceback bound to an exception object:

try:
    raise ValueError('Custom error message')
except Exception as err:
    tb = err.__traceback__
    print(tb)


Speaking your logger's language

Use logger level and format to dictate verbosity and mannerism:

logging.basicConfig(
    level=logging.DEBUG,
    format='%(asctime)s - %(levelname)s - %(message)s'
)
# Now, every log speaks in a nicer, more informative tone


Choosing your traceback gear

Select between traceback.print_exc(), logger.exception(), or traceback.format_exc(). It's like choosing between print, log, or store - Variety is spice of life, after all!