royerlab / arbol

arbol - Arborescent Printouts in Python

arbol | Python Package for Arborescent Printouts in Python

Do you have a script, a command line tool, or some workflow in Python that has lots of 'print' statements, and you can't make sense of it once it adds up to hundreds of lines on the console? Sounds familiar?
arbol organises your stdout prints in a hierarchy that follows the structure of your code. Use a simple context manager to define the hierarchy and the 'aprint' command instead of print, and voila. Finally, when the optional dependencies are installed, the printed tree and text are colored with an exquisitely crafted combination of colors, making it even more visually appealing.

If you are wondering, 'arbol' means 'tree' in spanish.

Why not use a more traditional Python logging? We have made the choice of sticking to a plain and simple scheme that matches the usage of 'print' statements.

Features

You can use 'aprint' replacement for the built-in 'print'. You create a new 'node' in the tree with the 'asection' contact manager. Moreover, arbol measures the elapsed time for each node of the tree and displays that conveniently. You have several configuration flags in the Arbol class to tune things. If yo want to capture print statements originating from 3rd party code, you can use the 'acapture' context manager to capture stdout (and stderr) -- tis is a bit experimental and is best used sparingly. The best documentation is simply the demo below...

Installation

Install with pip:

pip install arbol

Optional dependencies

If you want colors, install the ansicolors package:

pip install ansicolors

To make sure that you get the colors on all operating systems (particularly Windows), install the colorama package:

pip install colorama

Note: both colorama and ansicolors are optional -- arbol will work fine without it installed.

Example

Here is a simple and self-explanatory example:

from arbol import Arbol, aprint, section, asection, acapture

import arbol

# for colors, install the ansicolors package: 'pip install ansicolors',
# and for windows install the colorama package: 'pip install colorama'

# You can limit the tree depth:
Arbol.max_depth = 4

# use aprint (=arbol print) instead of the standard print
aprint('Test')

# You can decorate functions:
@section('function')
def fun(x):
    if x >= 0:
        with asection('recursive call to f'):
            aprint(f"f(x)+1={fun(x - 1)}")

# The context manager let's you go down one level in the tree
with asection('a section'):
    aprint('a line')
    aprint('another line')
    aprint('we are done \n or are we? \n someone gotta check!')

    with asection('a subsection'):
        aprint('another line')
        aprint('we are done')

    # works through function calls and the like...
    fun(2)

    # You can capture stdout if you want, usefull when a 3rd party library has printouts that you want to capture...
    with acapture():
        print("No escape is possible")
        aprint("Even this works...\n")
        # Don't push it.. sections will not work right now...

    # You can deactivate the elapsed time measurement and printing:
    Arbol.elapsed_time = False
    fun(100)

aprint('demo is finished...')

# You can also turn off all output with one switch:
Arbol.enable_output = False
aprint('you will not see that')

And this is how it looks like:

Roadmap

Some ideas we might consider from serious to highly speculative:

• More color styles to choose from
• Intercept stdout from C code so that printouts from libraries called from python are formatted too, unclear is that's possible.
• Generate tree automatically by inspecting stack?
• Interoperability with logging package?
• How to handle multiple threads/processes? Right now, the printouts get inter;leaved stochastically which is messy and incomprehensible. One idea is to capture all outputs from each thread, hold onto these outputs until the thread is done (perhaps via a dedicated 'holding' context manager), and ouput it all in order and separately. Probably doable.

Contributions

Pull requests highly welcome!

Authors

Loic A. Royer (@loicaroyer) Ahmet Can Solak (@_ahmetcansolak)