verbosify

Python decorator for turning on/off print statements in a function.

Why verbosify?

Because logging is for suckers! Just kidding. This is simply a quick and dirty decorator that is useful for debugging code without having to comment out print statements. It also allows you to rapidly add in on/off functionality for printing output (looking at you statsmodels ಠ_ಠ)

Usage

Simply decorate a function with @verbose to do the following:

• add the keyword-only argument verbose to the function
• update the docstring to include information about the verbose option
• update the function signature to show verbose with its default value (also adds completion for IPython)

The default behavior is to turn off printing

from verbosify import verbose

@verbose
def test(hello='world'):
    '''
    Test function to print an input.

    Parameters:
    -----------
    hello : <any>
        Any input.

    Returns:
    --------
    None
    '''
    print(hello)

Bringing up the help for test shows the modified docstring and arguments

help(test)
# prints
Help on function test in module __main__:

test(hello='world', *, verbose=False)
    Test function to print an input.

    Parameters:
    -----------
    hello : <any>
        Any input.

    Returns:
    --------
    None

    Option:
    -------
    verbose : bool
        Turns on/off print lines in the function.

Calling the function test prints nothing, unless verbose is set to True:

test()   # prints nothing
test(verbose=True)
# prints:
world

verbose can also be used as a factory to create decorators for turning print on/off by default

echo_off = verbose(default=False)  
echo_on = verbose(default=True)

# equivalent to @verbose or @verbose(default=False)
@echo_off
def test(x):
    print(x)
    
# equivalent to @verbose(default=True)
@echo_on
def test(x):
    print(x)

To do list:

• add argument to change the name of the keyword argument added to the decorated function (default: verbose)
• add argument to turn change the style of the appended docstring, possible styles:
  • 'none'
  • 'numpy'
  • 'sphinx'
  • 'google'
  • custom maybe?