Skip to content

Instantly share code, notes, and snippets.

@vincentwyshan
Forked from nealtodd/decorator.py
Last active February 26, 2024 15:30
Show Gist options
  • Select an option

  • Save vincentwyshan/62421a2ad7a16be07373f77dffdd982f to your computer and use it in GitHub Desktop.

Select an option

Save vincentwyshan/62421a2ad7a16be07373f77dffdd982f to your computer and use it in GitHub Desktop.
Python profiling decorator
"""
Python decorator for profiling a function using cProfile. Stats are output to wherever print statements would
normally go to, e.g. stdout or a log file. Can be used with Django view functions.
pstats' sort and print arguments can be supplied to the decorator. The default is to sort cumulatively and
list to first 10 lines.
See http://docs.python.org/library/profile.html#instant-user-s-manual for details on pstats sort and print arguments.
cProfile is a standard module in most python distributions. pstats is also in many but not all. If pstats doesn't
import then install the module via the python-profiler package. E.g. on apt-based systems:
sudo apt-get install python-profiler
"""
from cProfile import Profile
import pstats
def profile(sort_args=['cumulative'], print_args=[10]):
profiler = Profile()
def decorator(fn):
def inner(*args, **kwargs):
result = None
try:
result = profiler.runcall(fn, *args, **kwargs)
finally:
stats = pstats.Stats(profiler)
# stats.strip_dirs().sort_stats(*sort_args).print_stats(*print_args)
stats.sort_stats(*sort_args).print_stats(*print_args)
return result
return inner
return decorator
# Example without profiling arguments:
@profile():
def my_function(my_args):
pass
# Example with profiling arguments:
@profile(sort_args=['cumulative', 'name'], print_args=[.5, 'init']):
def my_function(my_args):
pass
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment