The idea is simple:

1. Read configuration from a YAML file
2. Locate the name of the log file
3. Append the date stamp to the name part (not the extension part)
4. Call logging.config.dictConfig and pass in the modified configuration dictionary

Here is the YAML file I use for this example, daily_log_file.yaml:

version: 1 loggers: default_logger: handlers: [consoleHandler, fileHandler] level: DEBUG handlers: consoleHandler: class: logging.StreamHandler level: DEBUG formatter: brief fileHandler: class: logging.FileHandler formatter: brief filename: '/tmp/daily_log_file.log' level: DEBUG formatters: brief: format: '%(levelname)8s %(message)s' 

Here is the script, daily_log_file.py:

import datetime import os import logging import logging.config import yaml def yaml_config(yaml_filename): global config_dict with open(yaml_filename) as f: config_dict = yaml.load(f) # Append the date stamp to the file name log_filename = config_dict['handlers']['fileHandler']['filename'] base, extension = os.path.splitext(log_filename) today = datetime.datetime.today() log_filename = '{}{}{}'.format( base, today.strftime('_%Y%m%d'), extension) config_dict['handlers']['fileHandler']['filename'] = log_filename # Apply the configuration logging.config.dictConfig(config_dict) if __name__ == '__main__': yaml_config('daily_log_file.yaml') logger = logging.getLogger('default_logger') logger.debug('debug message') logger.info('info message') 

Discussion

• yaml_config is where the action is. I first load the configuration dictionary from the YAML file, then patch up the file name
• To patch up the file name, I separate the base file name and extension
• I then get today's date and append the time stamp to the file name, then assign the result back to the configuration dictionary
• Finally, I called dictConfig to finish the job