Python Comprehensive Logging using YAML Configuration

logging.py

import os
import yaml
import logging.config
import logging
import coloredlogs

def setup_logging(default_path='logging.yaml', default_level=logging.INFO, env_key='LOG_CFG'):
    """
    | **@author:** Prathyush SP
    | Logging Setup
    """
    path = default_path
    value = os.getenv(env_key, None)
    if value:
        path = value
    if os.path.exists(path):
        with open(path, 'rt') as f:
            try:
                config = yaml.safe_load(f.read())
                logging.config.dictConfig(config)
                coloredlogs.install()
            except Exception as e:
                print(e)
                print('Error in Logging Configuration. Using default configs')
                logging.basicConfig(level=default_level)
                coloredlogs.install(level=default_level)
    else:
        logging.basicConfig(level=default_level)
        coloredlogs.install(level=default_level)
        print('Failed to load configuration file. Using default configs')

logging.yaml

version: 1
disable_existing_loggers: true



formatters:
    standard:
        format: "%(asctime)s - %(name)s - %(levelname)s - %(message)s"
    error:
        format: "%(levelname)s <PID %(process)d:%(processName)s> %(name)s.%(funcName)s(): %(message)s"

handlers:
    console:
        class: logging.StreamHandler
        level: DEBUG
        formatter: standard
        stream: ext://sys.stdout

    info_file_handler:
        class: logging.handlers.RotatingFileHandler
        level: INFO
        formatter: standard
        filename: /tmp/info.log
        maxBytes: 10485760 # 10MB
        backupCount: 20
        encoding: utf8

    error_file_handler:
        class: logging.handlers.RotatingFileHandler
        level: ERROR
        formatter: error
        filename: /tmp/errors.log
        maxBytes: 10485760 # 10MB
        backupCount: 20
        encoding: utf8

    debug_file_handler:
        class: logging.handlers.RotatingFileHandler
        level: DEBUG
        formatter: standard
        filename: /tmp/debug.log
        maxBytes: 10485760 # 10MB
        backupCount: 20
        encoding: utf8

    critical_file_handler:
        class: logging.handlers.RotatingFileHandler
        level: CRITICAL
        formatter: standard
        filename: /tmp/critical.log
        maxBytes: 10485760 # 10MB
        backupCount: 20
        encoding: utf8

    warn_file_handler:
        class: logging.handlers.RotatingFileHandler
        level: WARN
        formatter: standard
        filename: /tmp/warn.log
        maxBytes: 10485760 # 10MB
        backupCount: 20
        encoding: utf8

root:
    level: NOTSET
    handlers: [console]
    propogate: yes

loggers:
    <module>:
        level: INFO
        handlers: [console, info_file_handler, error_file_handler, critical_file_handler, debug_file_handler, warn_file_handler]
        propogate: no

    <module.x>:
        level: DEBUG
        handlers: [info_file_handler, error_file_handler, critical_file_handler, debug_file_handler, warn_file_handler]
        propogate: yes

I tried the code by @alexandertsema above, after adding the missing quote mark in the 'with open' line, and after simplifying by commenting out the filter code. Executing code below creates the info.log file, but it is empty. I wonder if something might be wrong with my python installation, which has trouble with all logging except the basicConfig, but seems to work fine for all other code. Suggestions?

setupLogging.py:

#!/usr/bin/env python3

import logging
import logging.config
import yaml

'''
class infoFilter(logging.Filter):
    def filter(self, rec):
        return rec.levelno == logging.INFO
'''

def setupLogging():
    with open('loggingConfig.yaml', 'rt') as file:
        config = yaml.safe_load(file.read())
        logging.config.dictConfig(config)

if __name__ == "__main__":
    setupLogging()
    logger = logging.getLogger(__name__)
    logger.info("logger set :D")

logging.yaml:

---

version: 1

disable_existing_loggers: False

#filters:
#    infoFilter:
#        (): setupLogging.infoFilter

handlers:
    info_file_handler:
        class: logging.handlers.TimedRotatingFileHandler
        level: INFO
        filename: info.log
        when: 'midnight'
        backupCount: 4
        utc: True
#        filters: [infoFilter]
        encoding: utf8

root:
    handlers: [info_file_handler]
...

My version of the code needs the line logger.setLevel(logging.DEBUG) after the getting the logger.

Refer for more details:
https://kingspp.github.io/design/2017/11/06/the-head-and-tail-of-logging.html

Thanks for this nice logging-config @kingspp. It helped a lot. And also big thanks to the filters @rochamorelli. This is exactly what I was searching!

Just one change to your code.
You set the filter in the yaml with:

filters:
  infoFilter:
    (): setupLogging.infoFilter

This didn't work for me because Python was complaining about that setupLogging is not a module. Since I'm creating the filter-class in a separate module logging.filters I changed it to the following:

filters:
  infoFilter:
    (): logging.filters.infoFilter

However, if you setup everything within a main-script in the root-folder of your program (also with defining the respective filter) it should be:

filters:
  infoFilter:
    (): __main__.infoFilter

I think your example was referring to the latter.

Hi all,
Thanks for the logging-config,
But I have some issue Executing code below creates all files, but they are empty.
Could you please help me?
Thanks

Hi all,
Thanks for the logging-config,
But I have some issue Executing code below creates all files, but they are empty.
Could you please help me?
Thanks

Could you add your code, please?

Ok, I've tried to use this method but for some reason DEBUG messages are not getting printed out in root logger.

main_file.py

import os
import yaml
import logging.config
import logging
import coloredlogs

import setup_logging
import child_module

setup_logging.setup_logging(default_level=logging.DEBUG)
logger = logging.getLogger(__name__)

logger.info("Here's the main file.")
child_module.some_function()

child_module.py

import logging
import setup_logging

setup_logging.setup_logging(default_level=logging.DEBUG)
logger = logging.getLogger(__name__)

logger.info("Imported child module")

def some_function():
    logger.debug("Debug inside the function")
    logger.info("Info inside the function")
    logger.warning("Warning inside the function")
    logger.error("Error inside the function")
    logger.critical("Critical inside the function")

And it does not display the logs inside the function of child module:

2020-09-27 22:01:45 child_module[24180] INFO Imported child module
2020-09-27 22:01:45 __main__[24180] INFO Here's the main file.

Ok, so I tinker the logging.yaml file and I enable other loggers: disable_existing_loggers: false
I'm able to get the functions INFO and higher logs but DEBUG is missing:

2020-09-27 22:02:45 child_module[22932] INFO Imported child module
2020-09-27 22:02:45 __main__[22932] INFO Here's the main file.
2020-09-27 22:02:45 child_module[22932] INFO Info inside the function
2020-09-27 22:02:45 child_module[22932] WARNING Warning inside the function
2020-09-27 22:02:45 child_module[22932] ERROR Error inside the function
2020-09-27 22:02:45 child_module[22932] CRITICAL Critical inside the function

Can someone tell me what's wrong, please?

FYI:

root:
    level: NOTSET
    handlers: [console]
    propogate: no

For some reason, one cannot set propagate to no in the root logger. It's always yes.

This can be verified easily by running the following print statement after logging.config.dictConfig(config) is made:

rl = logging.getLogger()
print(rl.propagate)
print(rl.getEffectiveLevel)
print(rl.handlers)

Python logging does not allow disabling propagation for root logger.
Ref: root logger section in dictConfig

What changes in configuration should I make to allow different Formatters for console, based on their level

formatters:
  humanDebug:
    (): flask_logify.formatters.RequestFormatter
    format: \x1b[91m"[%(asctime)s] [%(levelname)s] : %(message)s"\x1b[0m
  humanInfo:
    (): flask_logify.formatters.RequestFormatter
    format: \x1b[93m"[%(asctime)s] [%(levelname)s] : %(message)s"\x1b[0m

handlers:
  console:
    class: logging.StreamHandler
    stream: ext://sys.stdout

  consoleDebug:
    class: logging.StreamHandler
    stream: ext://sys.stdout
    formatter: humanDebug
    level: DEBUG

  consoleInfo:
    class: logging.StreamHandler
    stream: ext://sys.stdout
    formatter: humanInfo
    level: INFO

loggers:
  flask-development:
    handlers: [ consoleDebug,consoleInfo,consoleWarn]
    level: DEBUG
    propogate: yes

The current configuration would send out multiple messages, as a consoleDebug handler emits debug & above messages and so on and so forth. Can I limit a handler to specific level ? Not like DEBUG and above

Actually LOG_CFG is not needed, if we will use local search for logging.yml:

file = os.path.join(str(Path().absolute()), *__path__.split("/"), "logging.yaml")

I still struggle to make this work., though promising.

I did try to digest https://kingspp.github.io/design/2017/11/06/the-head-and-tail-of-logging.html which the OP has posted.

We have defined a function which can be used for configuring Root Logger, which stands atop logging hierarchy, now let see how we can use this function, Open __init__.py of main module. Why we need to use __init__.py will be discussed in upcoming posts, for the time being bear with me,

and the code therein,
`# init.py

import os
from logging_config_manager import setup_logging

setup_logging(default_path=os.path.join("/".join(file.split('/')[:-1]), 'config', 'logging_config.yaml'))`

these seem to indicate that if one calls teh setup_logging() each time at each module which one wish to log, it will treat each module as "root" hence all the lovely customisation wouldn't kick in at all, (which what I have seen playing at my side)

Does anyone had a true "working" example that one can share? please?

https://gist.github.com/glenfant/4358668
is a simple example on line. There you can see "logging" only been imported once.
but for a layered multi-module package, how would one possibly achieve that please?

any thoughts are much apprieciated please.

thanks

My version of the code needs the line logger.setLevel(logging.DEBUG) after the getting the logger.

i have the same issue of "missing the DEBUG" level. I have tried many many different ways
more importantly, here is my yaml file:

version: 1
disable_existing_loggers: True



formatters:
    standard:
        format: "%(asctime)s - %(name)s - %(levelname)s - %(message)s"
    error:
        format: "%(levelname)s <PID %(process)d:%(processName)s> %(name)s.%(funcName)s(): %(message)s"

filters:
    infoFilter:
        (): setup.logging_config_manager.infoFilter
    debugFilter:
        (): setup.logging_config_manager.debugFilter
    errorFilter:
        (): setup.logging_config_manager.errorFilter
    warningFilter:
        (): setup.logging_config_manager.warningFilter            
    criticalFilter:
        (): setup.logging_config_manager.criticalFilter

handlers:
    console:
        class: logging.StreamHandler
        level: INFO
        formatter: standard
        stream: ext://sys.stdout

    info_file_handler:
        class: logging.handlers.RotatingFileHandler
        level: INFO
        formatter: standard
        filename: /tmp/info.log
        maxBytes: 10485760 # 10MB
        backupCount: 20
        filters: [infoFilter]
        encoding: utf8

    error_file_handler:
        class: logging.handlers.RotatingFileHandler
        level: ERROR
        formatter: error
        filename: /tmp/errors.log
        maxBytes: 10485760 # 10MB
        backupCount: 20
        filters: [errorFilter]
        encoding: utf8

    debug_file_handler:
        class: logging.handlers.RotatingFileHandler
        level: DEBUG
        formatter: standard
        filename: /tmp/debug.log
        maxBytes: 10485760 # 10MB
        backupCount: 20
        encoding: utf8

    critical_file_handler:
        class: logging.handlers.RotatingFileHandler
        level: CRITICAL
        formatter: standard
        filename: /tmp/critical.log
        maxBytes: 10485760 # 10MB
        backupCount: 20
        filters: [criticalFilter]
        encoding: utf8

    warn_file_handler:
        class: logging.handlers.RotatingFileHandler
        level: WARN
        formatter: standard
        filename: /tmp/warn.log
        maxBytes: 10485760 # 10MB
        backupCount: 20
        filters: [warningFilter]
        encoding: utf8

root:
    level: DEBUG
    handlers: [console, info_file_handler]
    propagate: no

loggers:

    <__main__>:
        level: INFO
        handlers:  [ info_file_handler, debug_file_handler, error_file_handler, critical_file_handler, warn_file_handler]
        propagate: no


    <module1>:
        level: INFO
        handlers:  [ info_file_handler, debug_file_handler, error_file_handler, critical_file_handler, warn_file_handler]
        propagate: no

    <module1.x>:
        level: INFO
        handlers:  [error_file_handler]
        propagate: no   

so at "root" level, i should have DEBUG as starting point.
and I was trying to achieve that
for the main, module1, module1.x the logger will have different handlers. but those handlers do not seem kick in at all. in fact all that matters seems to be what i put in at "root" level. the logs are produced on console and for the Info.log file.

p.s., i have played around the propagte, disable_existing_loggers, and see no effects.
pps, i have done this logging_setup ONLY at the main level. as suggested from other stackoverflow sites; to avoid not loading the logging.config.dictConfig() multiple times at different module levels. (which is what Icmtcf has done in the previous post)

version: 1
disable_existing_loggers: no

formatters:
  standard:
    format: "%(asctime)s - %(name)s - %(levelname)s - %(message)s"
  error:
    format: "%(asctime)s - %(name)s - %(levelname)s <PID %(process)d:%(processName)s> %(name)s.%(funcName)s(): %(message)s"

handlers:
  info_file_handler:
    class: logging.handlers.RotatingFileHandler
    level: INFO
    formatter: standard
    filename: logs/info.log
    maxBytes: 10485760 # 10MB
    backupCount: 20
    encoding: utf8

  warn_file_handler:
    class: logging.handlers.RotatingFileHandler
    level: WARN
    formatter: standard
    filename: logs/warn.log
    maxBytes: 10485760 # 10MB
    backupCount: 20
    encoding: utf8

  error_file_handler:
    class: logging.handlers.RotatingFileHandler
    level: ERROR
    formatter: error
    filename: logs/errors.log
    maxBytes: 10485760 # 10MB
    backupCount: 20
    encoding: utf8

  critical_file_handler:
    class: logging.handlers.RotatingFileHandler
    level: CRITICAL
    formatter: standard
    filename: logs/critical.log
    maxBytes: 10485760 # 10MB
    backupCount: 20
    encoding: utf8

  debug_file_handler:
    class: logging.handlers.RotatingFileHandler
    level: DEBUG
    formatter: standard
    filename: logs/debug.log
    maxBytes: 10485760 # 10MB
    backupCount: 20
    encoding: utf8

  root_file_handler:
    class: logging.handlers.RotatingFileHandler
    level: DEBUG
    formatter: standard
    filename: logs/logs.log
    maxBytes: 10485760 # 10MB
    backupCount: 20
    encoding: utf8

  console:
    class: logging.StreamHandler
    level: DEBUG
    formatter: standard
    stream: ext://sys.stdout

  error_console:
    class: logging.StreamHandler
    level: ERROR
    formatter: error
    stream: ext://sys.stderr

root:
  level: DEBUG
  handlers: [console, error_console, root_file_handler]
  propagate: yes

loggers:
  main:
    level: DEBUG
    handlers: [info_file_handler, warn_file_handler, error_file_handler, critical_file_handler, debug_file_handler]
    propagate: no
  werkzeug:
    level: DEBUG
    handlers: [info_file_handler, warn_file_handler, error_file_handler, critical_file_handler, debug_file_handler]
    propagate: yes
  api.app_server:
    level: DEBUG
    handlers: [info_file_handler, warn_file_handler, error_file_handler, critical_file_handler, debug_file_handler]
    propagate: yes

Can you please give an example of how to write yaml file for aiologger?