snaplog

Zero-config structured logging for Python. One import, structured JSON output everywhere.

Zero dependencies. Auto-context. Function tracing. JSON by default.

Install

pip install snaplog

Quick Start

from snaplog import get_logger

log = get_logger()

log.info("Order processed", order_id="abc", total=49.99)
# → {"ts":"2026-03-24T21:08:00.123Z","level":"info","msg":"Order processed","order_id":"abc","total":49.99,"logger":"root"}

That's it. No config files. No handlers. No formatters. Just log.

Features

• 🚀 Zero config — get_logger() and go
• 📝 Structured JSON — every log is a JSON object
• 🎯 Auto-context — function name, file, line number captured automatically
• 🔍 Function tracing — @log.trace decorator adds timing
• 🏷️ Request context — with log.context(user_id=...) adds fields to all logs
• 👁️ Human-readable mode — for development
• 🔌 Stdlib integration — bridges to Python's logging module
• 🧵 Thread-safe — context propagates correctly
• 📦 Zero dependencies — pure Python

API

Basic Logging

from snaplog import get_logger

log = get_logger()
log.debug("Verbose details", step=1)
log.info("Something happened", user_id="u1")
log.warning("Careful!", retries=3)
log.error("Failed to connect", host="db1", error_code=503)
log.critical("System down", component="database")

Auto-Context

# Automatically includes function name, file, and line
def process_order(order_id):
    log.info("Processing", order_id=order_id)
    # → {..., "func":"process_order", "file":"orders.py", "line":42}

Function Tracing

@log.trace_fn
def fetch_user(user_id):
    return db.get(user_id)
# → {"level":"trace","msg":"enter: fetch_user","func":"fetch_user","args":{"user_id":"u1"}}
# → {"level":"trace","msg":"exit: fetch_user","func":"fetch_user","elapsed_ms":12.3,"result":"ok"}

Request Context

with log.context(request_id="r1", user_id="u42"):
    log.info("Processing request")  # includes request_id + user_id
    call_service()
    log.info("Done")  # still includes them
# Context automatically removed after `with` block

Human-Readable Output

from snaplog import get_logger, HumanFormatter

log = get_logger(formatter=HumanFormatter())
log.info("Hello", name="world")
# → 2026-03-24 21:08:00 [INFO] Hello  name=world

Min Level

log = get_logger(min_level="warning")  # debug, info suppressed

Stdlib Integration

import logging
from snaplog import SnaplogHandler

handler = SnaplogHandler()
logging.root.addHandler(handler)
logging.info("This goes through snaplog too")

Custom Output

# Write to a file
import sys
log = get_logger(output=open("app.log", "a"))

# Custom output function
def my_output(record_dict):
    requests.post("https://logs.example.com", json=record_dict)

log = get_logger(output=my_output)

Output Format

{
  "ts": "2026-03-24T21:08:00.123Z",
  "level": "info",
  "msg": "Order processed",
  "logger": "root",
  "func": "process_order",
  "file": "orders.py",
  "line": 42,
  "order_id": "abc",
  "total": 49.99
}

License

MIT