(My) Notes about Logging in Python

Photo by Cathryn Lavery on Unsplash

The Problem:

❔I want to get more information when running an application and keep a log of potential errors. All posts or notes I found are long, complicated, or too advanced.

The solution:

🍄 Level up as a developer and start using Logging instead of print statements and make short simple notes.

The Implementation:

🌲Use the logging module in python, to replace the print statements and create a file handler to create log files. Present two examples: A Basic one oriented to quick implementation, and an Advanced one for more complex projects.

Why should I use it:

The module Logging will allow me to get an overview of errors or the data flow in specific code. For small projects, a simple print will help to check results or errors in specific statements. But in a more complex project, where I have several functions and modules, logging is a better and more organized option.

The configuration of the logger can get complicated however it is worth it.

The Basics:

The Logging module is a built-in module, so installation is not necessary.

I only need to import it.

import logging

There are several elements in the logging, but the most important are:

• Logging levels
• Handler
• Formatter

Logging Levels

Logging levels will allow me to log more specific things, such as; debugging statements, information, or report errors.

🔥 The default level is set to “WARNING”. If the logging level is not configured, the only messages logged are those higher than WARNING, in other words, ERROR and CRITICAL.

Example of a basic logging

This example will show the simple logging configuration. There is no configuration for a handler or a formatter. However, the implementation will allow me to decide whether or not I want to log the information in a file and in what format I want the information.

Just with the logging level, the default output will be the console.

Basic configuration — Logging level

Adding a destination file

Basic configuration — File name and Logging level

Changing the format of the output

Basic configuration — File name, Logging level, and Format for message

In the above example, I configure the logs to display the time and a message, But these are not the only attributes I can use. python documentation provides an extensive list of LogRecord attributes here:

logging - Logging facility for Python - Python 3.10.4 documentation

The complete code of the basic example will be:

A basic example of a logger to a file

Advanced Logging

Photo by Alex Knight on Unsplash

I have just one logger in the following example — a good option if I deal with one script at a time. But for a more complex app, I will need more loggers and probably more output.

To achieve a more functional logger, I can make use of the elements mentioned previously

• Logging levels
• Handler
• Formatter

The following code can be duplicated in each script to have a logger in each one of them

Creating a logger for a specific script

I start by creating a variable:

Advanced example Logger

The argument __name__ is a deeper topic that I won't cover here. but a bad explanation or rather a not accurate one will say; “this will help the lower to know in which module and function it is been call”.

Logging Levels

To add the logging level to the logger, I use the function setLevel().

Advanced example Logger — Set Level

Handler

In this case refer to which output I will use, the console or a document.

Advanced example Logger — Handler

I created a file handler but it is not linked to the logger yet, I need to add it.

Advanced example Logger — Add Handler

Be aware that I haven’t changed the format of the output, I just define what will be the output for the logger.

To add a stream handler or an output to the console, I can use StreamHandler().

Advanced example Logger — Add Handlers

Formatter

Now with the logger added I can think in the format I want. To modify the format I can use setFormatter().

Advanced example Logger — Formatter

The extra argument datefmt allows me to modify how the time is displayed. Similar to Handler, I have to register it with the logger. setFormatter() is used to add the format to the logger.

Advanced example Logger — Set Formatter

Now, all together:

Advanced example Logger

🔥 Every time I need to call the logger, I will need to use the variable I create for it rather than the logging.

Example:

logger.info(f'Starting script')

Notes:

When using try/except blocks is a good idea to log the traceback when the action fails by using the function .exception() with the logger.

Victor Fernandez

• Medium
• My LinkedIn