The library helps you to store logs in JSON format. Why is it important? Well, it facilitates integration with Logstash.

Usage example:

import logging
import json_log_formatter
formatter = json_log_formatter.JSONFormatter()
json_handler = logging.FileHandler(filename='/var/log/my-log.json')
json_handler.setFormatter(formatter)
logger = logging.getLogger('my_json')
logger.addHandler(json_handler)
logger.setLevel(logging.INFO)
logger.info('Sign up', extra={'referral_code': '52d6ce'})
try:
    raise ValueError('something wrong')
except ValueError:
    logger.error('Request failed', exc_info=True)

The log file will contain the following log record (inline).

{
    "message": "Sign up",
    "time": "2015-09-01T06:06:26.524448+00:00",
    "referral_code": "52d6ce"
}
{
    "message": "Request failed",
    "time": "2015-09-01T06:06:26.524449+00:00",
    "exc_info": "Traceback (most recent call last): ..."
}

If you use a log collection and analysis system, you might need to include the built-in log record attributes with VerboseJSONFormatter.

json_handler.setFormatter(json_log_formatter.VerboseJSONFormatter())
logger.error('An error has occured')

{
    "filename": "tests.py",
    "funcName": "test_file_name_is_testspy",
    "levelname": "ERROR",
    "lineno": 276,
    "module": "tests",
    "name": "my_json",
    "pathname": "/Users/bob/json-log-formatter/tests.py",
    "process": 3081,
    "processName": "MainProcess",
    "stack_info": null,
    "thread": 4664270272,
    "threadName": "MainThread",
    "message": "An error has occured",
    "time": "2021-07-04T21:05:42.767726+00:00"
}

If you need to flatten complex objects as strings, use FlatJSONFormatter.

json_handler.setFormatter(json_log_formatter.FlatJSONFormatter())
logger.error('An error has occured')
logger.info('Sign up', extra={'request': WSGIRequest({
    'PATH_INFO': 'bogus',
    'REQUEST_METHOD': 'bogus',
    'CONTENT_TYPE': 'text/html; charset=utf8',
    'wsgi.input': BytesIO(b''),
})})

{
    "message": "Sign up",
    "time": "2024-10-01T00:59:29.332888+00:00",
    "request": "<WSGIRequest: BOGUS '/bogus'>"
}

JSON libraries

You can use ujson or simplejson instead of built-in json library.

import json_log_formatter
import ujson
formatter = json_log_formatter.JSONFormatter()
formatter.json_lib = ujson

Django integration

Here is an example of how the JSON formatter can be used with Django.

LOGGING['formatters']['json'] = {
    '()': 'json_log_formatter.JSONFormatter',
}
LOGGING['handlers']['json_file'] = {
    'level': 'INFO',
    'class': 'logging.FileHandler',
    'filename': '/var/log/my-log.json',
    'formatter': 'json',
}
LOGGING['loggers']['my_json'] = {
    'handlers': ['json_file'],
    'level': 'INFO',
}

Let's try to log something.

import logging
logger = logging.getLogger('my_json')
logger.info('Sign up', extra={'referral_code': '52d6ce'})

Custom formatter

You will likely need a custom log formatter. For instance, you want to log a user ID, an IP address and time as django.utils.timezone.now(). To do so you should override JSONFormatter.json_record().

class CustomisedJSONFormatter(json_log_formatter.JSONFormatter):
    def json_record(self, message: str, extra: dict, record: logging.LogRecord) -> dict:
        extra['message'] = message
        extra['user_id'] = current_user_id()
        extra['ip'] = current_ip()
        # Include builtins
        extra['level'] = record.levelname
        extra['name'] = record.name
        if 'time' not in extra:
            extra['time'] = django.utils.timezone.now()
        if record.exc_info:
            extra['exc_info'] = self.formatException(record.exc_info)
        return extra

Let's say you want datetime to be serialized as a timestamp. Override mutate_json_record to convert it.

class CustomisedJSONFormatter(json_log_formatter.JSONFormatter):
    def mutate_json_record(self, json_record):
        for attr_name, attr in json_record.items():
            if isinstance(attr, datetime):
                json_record[attr_name] = attr.timestamp()
        return json_record

Tests

$ pip install -r requirements.txt
$ tox