Don't fight with logging ...

Easy logging configuration based on environment variables.

Features:

• Set logging level using environment variable LOG_LEVEL (defaults to 'INFO')
• Set which loggers to enable using environment variable LOGGERS (defaults to '', everything)
• Always output to stdout
• Optional JSON formatter
• Completely disable logging setting LOG_LEVEL=DISABLED

Requirements:

• Python 3.5 and beyond

Install:

pip install belogging

# my_script.py

import belogging
belogging.load()
# ^^ this call is optional, only useful for customization
# For example, to enable JSON output: belogging.load(json=True)

# belogging.getLogger is just a sugar to logging.getLogger, you can
# use logging.getLogger as usual (and recommended).
logger = belogging.getLogger('foobar')
logger.debug('test 1')
logger.info('test 2')

Executing:

# selecting LOG_LEVEL
$ LOG_LEVEL=DEBUG python my_script.py
# level=DEBUG message=test 1
# level=INFO message=test 2

# selecting LOGGERS
$ LOGGERS=foobar python my_script.py
# Both messages

# Both
$ LOGGERS=foobar LOG_LEVEL=INFO my_script.py
# only level=INFO message=test 2

Applications should call `belogging.load()` upon initialization. The first `__init__.py` would be a good candidate, but anything before any call to `logging` module will be fine.

In your projects `settings.py`:

import belogging
# Disable django logging setup
LOGGING_CONFIG = None
belogging.load()

Inside your code, just use `logging.getLogger()` as usual.

$ export LOG_LEVEL=WARNING
$ ./manage.py runserver
# It will output only logging messages with severity > WARNING

Logging follows a hierarchy, so you easily select or skip some logging messages:

$ export LOGGERS=my_app.critical_a,my_app.critical_c,my_lib
$ ./my-app.py
# "my_app.critical_b messages" will be skipped
# all messages from my_lib will show up