philiprehberger-json-logger

Drop-in structured JSON logging for Python using the stdlib logging module.

Installation

pip install philiprehberger-json-logger

Usage

Quick setup

from philiprehberger_json_logger import setup

setup(level="DEBUG")

import logging
logging.info("Server started", extra={"port": 8080})
# {"timestamp": "2026-03-13T12:00:00+00:00", "level": "INFO", "message": "Server started", "logger": "root", "module": "app", "line": 5, "port": 8080}

Named logger

import logging
from philiprehberger_json_logger import setup

logger = logging.getLogger("myapp")
setup(level="INFO", logger=logger)

logger.warning("Disk usage high", extra={"usage_pct": 91.3})

Static extra fields

from philiprehberger_json_logger import setup

setup(level="INFO", extra_fields={"service": "api", "env": "production"})

Every log entry will include "service": "api" and "env": "production".

Field redaction

from philiprehberger_json_logger import setup

setup(level="INFO", redact_fields={"password", "token", "secret"})

import logging
logging.info("Login", extra={"user": "alice", "password": "s3cret"})
# password field will appear as "***"

Redaction works recursively on nested dicts.

Scoped context

import logging
from philiprehberger_json_logger import setup, log_context

setup(level="INFO")

with log_context(request_id="abc-123", user="alice"):
    logging.info("Processing request")
    # log entry includes request_id and user fields

    with log_context(step="validation"):
        logging.info("Validating input")
        # log entry includes request_id, user, and step fields

Custom handler with JsonFormatter

import logging
from philiprehberger_json_logger import JsonFormatter

handler = logging.FileHandler("app.log")
handler.setFormatter(JsonFormatter(extra_fields={"service": "worker"}, redact_fields={"token"}))

logger = logging.getLogger("worker")
logger.addHandler(handler)
logger.setLevel(logging.INFO)

API

Name	Description
JsonFormatter(*, extra_fields=None, redact_fields=None)	Logging formatter that outputs JSON lines. extra_fields is a dict of static fields merged into every entry. redact_fields is a set of field names replaced with "***".
setup(level="INFO", *, extra_fields=None, redact_fields=None, logger=None)	Configure a logger with JSON output. Defaults to the root logger. Clears existing handlers.
log_context(**kwargs)	Context manager that injects fields into all log entries within the block. Supports nesting.

JSON output fields

Field	Description
timestamp	ISO 8601 UTC timestamp
level	Log level name (DEBUG, INFO, WARNING, ERROR, CRITICAL)
message	Formatted log message
logger	Logger name
module	Source module name
line	Source line number
exception	Full traceback string (only present when logging an exception)

Any extra={} kwargs passed to the log call are merged into the top-level JSON object. Static extra_fields from the formatter and active log_context() fields are also merged.

Development

pip install -e .
python -m pytest tests/ -v

Support

If you find this project useful:

⭐ Star the repo

🐛 Report issues

💡 Suggest features

❤️ Sponsor development

🌐 All Open Source Projects

💻 GitHub Profile

🔗 LinkedIn Profile

License

MIT