d-e-f-e-a-t / cyjson

Fast JSON Parser for Python ( in Cython )

cyjson

Fast and efficient JSON Parser for Python.

This is a wrapper class written in Cython. The Underlaying library is cJSON by Dave Gamble written in C.

import cyj

parser = cyj.cyj()

#loading data into cyjson and parsing it internally
#returns True on success and False on failure
parser << '{"ab":"test","c":[1,2,{"x":["y","f","w",[6,7,8,10]]},4],"s":{"y":["llüöll",1,false,true]}}'

#information such as types, keywords and item size
print parser.root_info()

#getting information on elements, raises cyJSON_* exception on error
print parser >> ('c',2)

#converting items into python objects
for i in parser.root_info()["keys"]:
	print parser( ( str(i), ) )

loading

parser << "JSON" # will be converted -> .replace("\'","\"")

converting to python object

This is not efficient, because it converts the whole structure into python objects.

parser.get_root()

retrieving info

Instead of converting to python object, we can collect some information and extract the target value directly from C layer which is faster and more efficient.

>>> parser.root_info()
{'keys': [u'ab', u'c', u's'], 'types': ['str', 'list', 'dict'], 'size': 3}
>>> parser >> ('c',)
{'keys': [], 'types': ['num', 'num', 'dict', 'num'], 'size': 4}

get value

>>> parser( ('c',1) )
2
>>> parser( ('c',2 ) )
{'x': [u'y', u'f', u'w', [6, 7, 8, 10]]}

#install git clone --recursive https://github.com/mitghi/cyjson/ cd ./cyjson/cJSON; make ; make install ; cd .. python setup.py build_ext --inplace

#requirements Cython 0.23