logry 0.1.0

A fast, flexible Python logging library with sync and async support.

logry

Python logging library with native async support, multiple independent loggers, and zero external dependencies.

Installation

pip install logry

Requires Python 3.8 or higher.

---

Quick Start

from logry import log

log.info("Server started")
log.success("Connected to database")
log.warning("Memory usage at 80%")
log.error("Connection dropped")
log.debug("user_id=123")

Output:

[2026-03-09 10:23:45] INFO    | Server started
[2026-03-09 10:23:46] SUCCESS | Connected to database
[2026-03-09 10:23:47] WARNING | Memory usage at 80%
[2026-03-09 10:23:48] ERROR   | Connection dropped
[2026-03-09 10:23:49] DEBUG   | user_id=123

---

Log Levels

Level	Method	Color
DEBUG	log.debug()	Grey
INFO	log.info()	Blue
SUCCESS	log.success()	Green
WARNING	log.warning()	Yellow
ERROR	log.error()	Red

---

Usage

Global logger

The simplest way to use logry is via the built-in global log instance, which is ready to use immediately after import.

from logry import log

log.info("Application initialized")
log.error("Unexpected error")

Named loggers

Each Logger instance is fully independent and can be configured separately.

from logry import Logger

db  = Logger("database")
api = Logger("api")
bot = Logger("bot")

db.info("Connected to PostgreSQL")
api.error("Request timeout")
bot.success("Polling started")

Output formats

Three formats are available: pretty (default), minimal, and json.

pretty outputs colored lines with a timestamp, suited for development. minimal outputs level and message only. json outputs structured single-line JSON, suited for production log collection.

log = Logger("app", format="json")
log.info("Structured output")

{"time": "2026-03-09 10:23:45", "level": "INFO", "name": "app", "message": "Structured output"}

File output

log = Logger("app")
log.set_file("app.log")

With automatic rotation:

log.set_file("app.log", max_size="10MB", backup=3)

max_size accepts values in KB or MB. When the file reaches the specified size, it is rotated and up to backup copies are retained.

Custom colors

log.set_colors({
    "debug":   "cyan",
    "info":    "blue",
    "success": "green",
    "warning": "yellow",
    "error":   "red",
})

Available colors: black, red, green, yellow, blue, magenta, cyan, white, grey.

Full configuration

All options can be set at construction time or updated at runtime.

log = Logger(
    name="app",
    level="DEBUG",
    save="app.log",
    format="pretty",
    colors=True,
    async_mode=False,
)

log.set_level("WARNING")
log.set_format("json")
log.set_file("app.log", max_size="5MB", backup=5)

Async support

When async_mode=True, log calls become coroutines and must be awaited. Output is written through an asyncio.Queue worker, keeping the event loop unblocked.

import asyncio
from logry import Logger

log = Logger("app", async_mode=True)

async def main():
    await log.ainfo("Async log entry")
    await log.aerror("Async error")

asyncio.run(main())

---

Benchmarks

Tested on Python 3.12 with 100,000 filtered log entries. Filtered means the log level of the call is below the logger's active level, so no output is produced — the most common scenario in production.

Library	Time	Messages/sec
stdlib logging	0.085s	1,176,000
logry	0.031s	3,225,000

logry avoids formatting the message string entirely when the log level is inactive (lazy formatting), which accounts for the difference.

---

License

MIT © riokzy