python decorator is used in Logging, type checking, exception handling without making the functional code dirty
logging
def log_decorator(log_name):
def log_this(function):
logger = logging.getLogger(log_name)
def new_function(*args,**kwargs):
logger.debug(f"{function.__name__} - {args} - {kwargs}")
output = function(*args,**kwargs)
logger.debug(f"{function.__name__} returned: {output}")
return output
return new_function
return log_this
type checking
def accepts(*types):
def check_accepts(function):
assert len(types) == function.__code__.co_argcount,\
"Number of typed inputs must match the function inputs"
def new_function(*args, **kwargs):
for (a, t) in zip(args, types):
assert isinstance(a, t), \
"arg %r does not match %s" % (a,t)
return function(*args, **kwargs)
return new_function
return check_accepts
error handling
def error_handling(api_function):
def trial(*args, num_retries=0, **kwargs):
try:
return api_function(*args, **kwargs)
except api.error.RateLimitError:
if num_retries > MAX_RETRIES:
raise RuntimeError("Too many retries")
else:
msg = f"rate limit reached. Waiting {API_WAIT_TIME} minutes ..."
time.sleep(API_WAIT_TIME * 60)
return trial(*args, num_retries=num_retries + 1, **kwargs)
return trial
