char

==== char

.. image:: https://img.shields.io/github/last-commit/stas-prokopiev/char :target: https://img.shields.io/github/last-commit/stas-prokopiev/char :alt: GitHub last commit

.. image:: https://img.shields.io/github/license/stas-prokopiev/char :target: https://github.com/stas-prokopiev/char/blob/master/LICENSE.txt :alt: GitHub license

.. image:: https://readthedocs.org/projects/char/badge/?version=latest :target: https://char.readthedocs.io/en/latest/?badge=latest :alt: Documentation Status

.. image:: https://travis-ci.org/stas-prokopiev/char.svg?branch=master :target: https://travis-ci.org/stas-prokopiev/char

.. image:: https://img.shields.io/pypi/v/char :target: https://img.shields.io/pypi/v/char :alt: PyPI

.. image:: https://img.shields.io/pypi/pyversions/char :target: https://img.shields.io/pypi/pyversions/char :alt: PyPI - Python Version

.. contents:: Table of Contents

Overview.

char stands for: check of arguments.

This library gives to user ability to check types of function arguments via one decorator.

| If your team or you have some agreements how to name variables with defined types | Or if you are ready to use mine (which is derived from Hungarian notation (it will be described bellow)) | then type checking will be simple and pleasant.

P.S. I like to use variables notation because it really improves readability of the code.

Example

| Let's say that you've defined a function func(int_x) and you want to check | if value that is given to variable int_x has type int | Usually you would have to check it by adding one more line at the start of the function | somehow like this: isinstance(int_x, int) | With this decorator this can be done for you automatically.

.. code-block:: python

from char import char
# OR from char import check_types_of_arguments  # They are equivalent

@char
def func(int_x):
    pass

| If you try to call function with wrong types of arguments like: func("pewpew") | then you'll get a nice exception with description:

.. code-block:: bash

ArgumentTypeError: Incorrect type of variable was given to function: func
---> For variable: int_x
---> Were given value: pewpew
---> With type: <class 'str'>
---> Instead of: <class 'int'>

Differences from python type hinting

| The main difference from the internal python type hinting is that type hinting do static analysis of the code | and do not do any checks during the runtime. | So actually it can't protect a user from giving to some type hinted function arguments with the wrong types | so even with type hinting you still have to write isinstance type checks.

Additional difference is that this package supports python >=2.7 when type hinting is available only since python 3.5

In case if you don't want to use any notation of variables and want to use the type hinting then you can try the library typeguard.

Installation via pip:

.. code-block:: bash

pip install char

Usage with default settings

Default prefixes

Here will be a list of name prefixes and which type the variable is expected to be

If variable name doesn't start with ony of the given prefixes then variable type won't be checked.

#. "any_" - object #. "bool_" - bool #. "b_" - bool #. "is_" - bool #. "has_" - bool #. "str_" - str #. "bytes_" - bytes #. "int_" - int #. "i_" - int #. "float_" - float #. "f_" - float #. "list_" - list #. "l_" - list #. "dict_" - dict #. "d_" - dict #. "set_" - set #. "tuple_" - tuple #. "t_" - tuple

Example

.. code-block:: python

from char import char

@char
def oh_my_god(
        int_arg,
        float_arg,
        list_arg,
        undef_arg,
        d_kwarg=None,
        i_kwarg=0,
        is_kwarg=False
):
    pass

oh_my_god(0, 0.0, [], 1)  # Will PASS
oh_my_god(0, 0.0, None, "text")  # Will PASS
oh_my_god(0, 0.0, {}, "text")  # Will FAIL and raise an ArgumentTypeError
oh_my_god(0, 0.0, [], Exception, d_kwarg={0: 1})  # Will PASS
oh_my_god(0, 0.0, [], object, is_kwarg=0)  # Will FAIL and raise an ArgumentTypeError

Usage with user defined settings

Decorator arguments

#. bool_is_to_skip_None_value=True: Flag what to do with None values, by default None values won't be checked. #. dict_tuple_types_by_prefix_to_update_default: dictionary, which prefixes to add to the default ones #. dict_tuple_types_by_prefix: dictionary, which prefixes to use instead of default ones

| Your can use any combination of given arguments for the decorator. | For simplicity will be shown usage of every argument separately.

Decorator argument: bool_is_to_skip_None_value

.. code-block:: python

@char
def func_with_default_decorator(dict_x):
    pass

@char(bool_is_to_skip_None_value=False)
def func_with_custom_decorator(dict_x):
    pass

func_with_default_decorator(None)  # Will PASS
func_with_custom_decorator(None)  # Will FAIL and raise an ArgumentTypeError

Decorator argument: dict_tuple_types_by_prefix_to_update_default

.. code-block:: python

@char(dict_tuple_types_by_prefix_to_update_default={"num_": (int, float, bool)})
def very_complex_function(num_x, str_y=""):
    pass

very_complex_function(0, "hihi")  # Will PASS
very_complex_function(0.5, "heyhey")  # Will PASS
very_complex_function(True)  # Will PASS
very_complex_function("True")  # Will FAIL and raise an ArgumentTypeError

Decorator argument: dict_tuple_types_by_prefix

.. code-block:: python

@char(dict_tuple_types_by_prefix={"exception": (BaseException)})
def function_with_only_one_check(int_x, exception_y=None):
    pass

function_with_only_one_check(0, Exception)  # Will PASS
function_with_only_one_check(0.5, TypeError)  # Will PASS because first variable won't be checked
function_with_only_one_check(0.5, "ERROR")  # Will FAIL and raise an ArgumentTypeError

Links

* `PYPI <https://pypi.org/project/char/>`_
* `readthedocs <https://char.readthedocs.io/en/latest/>`_
* `GitHub <https://github.com/stas-prokopiev/char>`_

Project local Links

* `CHANGELOG <https://github.com/stas-prokopiev/char/blob/master/CHANGELOG.rst>`_.
* `CONTRIBUTING <https://github.com/stas-prokopiev/char/blob/master/CONTRIBUTING.rst>`_.

Contacts

* Email: [e-mail]
* `vk.com <https://vk.com/stas.prokopyev>`_
* `Facebook <https://www.facebook.com/profile.php?id=100009380530321>`_

License

This project is licensed under the MIT License.