Library to enrich structured logs

pip install log-enricher==1.0.0



CircleCI PyPI Downloads PyPI Version License

This is a log enricher, useful for adding custom fields to log records.

This was developed at GRID for use with our python backend services and intended to emit structured logs.


pip install log-enricher


log-enricher.initialize_logging(...) configures the logging library and takes in enrichers, a list of functions that return a dictionary. When a log message is sent, the enrichers are run automatically and their output is added to the log message, if structured logging is enabled.

Furthermore, initialize_logging() takes a list of loggers to use, a switch to control structured_logs (JSON logs, default on), and a log_level setting.

Logs will be output in a structured JSON format by default - if structured_logs is True - or in a plain, console-friendly format if structured_logs is False.

config example

import os

from log_enricher import initialize_logging, Level
from log_enricher.enrichers import Enricher

class UserContextEnricher(Enricher):
    def __call__(self) -> Dict[str, Any]:
        user_context = get_user_context()
        return {"username": user_context.get("username")}

extra_log_properties = {
    "app_version": Config.APP_VERSION, "release_stage": Config.RELEASE_STAGE

def main():
        loggers=["uvicorn", "sqlalchemy"],
        structured_logs=os.environ.get("STRUCTURED_LOGS", True),
        enrichers=[UserContextEnricher(), lambda: extra_log_properties],


To build a log enricher, make a subclass of Enricher, or Callable, and implement __call__(). Any method returning a dict can be used to enrich log records. See log_enricher/ The key-value pairs in the dict are added as attribute-value pairs to the log record. Of course, any method calls in the enrichers need to work in any subsequent context the logging system is called.