A Zoo for decorators
MIT License
A zoo for decorators
There are many great decorators out there that we use everyday. Why don't collect few of them?
I found myself implementing over and over in different projects. The hope is to gather them here and use this codebase.
This codebase is experimental and is working for my use cases. It is very probable that there are cases not covered and for which it breaks (badly). If you find them, please feel free to open an issue in the issue page of the repo.
In short a python decorator is a way to modify or enhance the behavior of a function or a class without actually modifying the source code of the function or class.
Decorators are implemented as functions (or classes) that take a function or a class as input and return a new function or class that has some additional functionality.
To have a more in-depth explanation you can check the decorators docs section.
deczoo is published as a Python package on pypi, and it can be installed with pip, or directly from source using git, or with a local clone:
pip (suggested):
python -m pip install deczoo
pip + source/git:
python -m pip install git+https://github.com/FBruzzesi/deczoo.git
local clone:
git clone https://github.com/FBruzzesi/deczoo.git
cd deczoo
python -m pip install .
As of now, the library has no additional required dependencies, however:
@memory_limit
and @timeout
)chime
to use @chime_on_end
)The idea is kind of simple: each function in the library is a (function) decorator with a specific objective in mind.
from deczoo import log
@log # equivalent to @log(log_time=True, log_args=True, log_error=True, logging_fn=print)
def custom_add(a, b, *args):
"""Adds all arguments together"""
return sum([a, b, *args])
_ = custom_add(1, 2, 3, 4)
# custom_add args=(a=1, b=2, args=(3, 4)) time=0:00:00.000062
_ = custom_add(1, "a", 2)
# custom_add args=(a=1, b=a, args=(2,)) time=0:00:00.000064 Failed with error: unsupported
# operand type(s) for +: 'int' and 'str'
from deczoo import shape_tracker
@shape_tracker(shape_in=True, shape_out=True, shape_delta=True, raise_if_empty=True)
def tracked_vstack(a: np.ndarray, b: np.ndarray) -> np.ndarray:
return np.vstack([a, b])
_ = tracked_vstack(np.ones((1, 2)), np.ones((10, 2)))
# Input: `a` has shape (1, 2)
# Output: result has shape (11, 2)
# Shape delta: (-10, 0)
The library implements the following decorators:
call_counter
: tracks how many times a function has been called.catch
: wraps a function in a try-except block, returning a custom value, or raising a custom exception.check_args
: checks that function arguments satisfy its "rule".chime_on_end
: notify with chime sound on function end (success or error).log
: tracks function time taken, arguments and errors, such logs can be written to a file.timer
: tracks function time taken.memory_limit
: sets a memory limit while running the function.notify_on_end
: notifies when function finished running with a custom notifier.raise_if
: raises a custom exception if a condition is met.retry
: wraps a function with a "retry" block.shape_tracker
: tracks the shape of a dataframe/array-like object, in input and/or output.multi_shape_tracker
: tracks the shapes of input(s) and/or output(s) of a function.timeout
: sets a time limit for the function, terminates the process if it hasn't finished within such time limit.Please refer to the api page to see a basic example for each decorator.
Please read the Contributing guidelines in the documentation site.
The project has a MIT Licence