strrapidjson is a fork of the pyrapidjson library (version 0.5.1), which in turn is a Python 2.7+ wrapper for rapidjson. For Python 3+ python-rapidjson is another alternative.
Main design goal of this library is to avoid using unicode() conversions in Python 2, sticking to (usually UTF8-encoded) str() instead.
- When decoding, library always creates str() objects
- When encoding, ensure_ascii=False is applied by default
There are performance advantages to avoiding conversion to unicode() objects when that is not needed.
Performance-wise, strrapidjson is faster than the standard library json module and comparable to ujson and simplejson packages.
from pip:
$ pip install strrapidjson
from pip+Github:
$ pip install git+https://github.com/aarki/strrapidjson
clone and run locally (rexcursive to pull rapidjson submodule):
$ git clone --recursive https://github.com/aarki/strrapidjson.git $ cd strrapidjson $ python setup.py install
Python2.7. Not tested in Python3+ after forking from pyrapidjson
Drop-in replacement for Python json module although function arguments to dump/load are not supported.
Example:
>>> import strrapidjson
>>> strrapidjson.loads('[1, 2, {"test": "hoge"}]')
>>> [1, 2, {'test': 'hoge'}]
>>> strrapidjson.dumps([1, 2, {"foo": "bar"}])
'[1,2,{"foo":"bar"}]'
>>>