python wrapper around the fast C++ logger called spdlog
Python wrapper (pybind11) around the C++ spdlog logging library.
Why choose spdlog?
https://kjellkod.wordpress.com/2015/06/30/the-worlds-fastest-logger-vs-g3log/
Try running tests/spdlog_vs_logging.py and see what results you get on your system.
How many microseconds it takes on average to complete a log function (info(), debug() etc) using a FileLogger. On reasonable sized log messages spdlog takes 4% (async mode enabled) and 6% (sync mode) of the time it would take to complete using the standard logging module.
Async mode with 8MB queue with blocking mode.
| msg len (bytes) | spdlog sync (microsec) | spdlog async (microsec) | logging (microsec) |
|---|---|---|---|
| 10 | 1.2 | 0.87 | 24.6 |
| 100 | 1.2 | 1.03 | 24.6 |
| 300 | 1.5 | 1.07 | 24.9 |
| 1000 | 2.4 | 1.16 | 26.8 |
| 5000 | 6.2 | 2.31 | 31.7 |
| 20000 | 15.3 | 7.51 | 48.0 |
pip install spdlogwill get a distribution from pypi.org
or
- from github:
pip install pybind11 - if missing
git clone https://github.com/bodgergely/spdlog-python.git
cd spdlog-python
git submodule update --init --recursive
python setup.py install./python
import spdlog as spd
logger = spd.FileLogger('fast_logger', '/tmp/spdlog_example.log')
logger.set_level(spd.LogLevel.INFO)
logger.info('Hello World!')
logger.debug('I am not so important.')To run the speed test:
python ./tests/spdlog_vs_logging.py