Cacher is the code snippet organizer for pro developers

#!/usr/bin/env python
# -*- coding: utf-8 -*-

"""Which json is the best json."""

import json
import simplejson as sjson
import ujson
import cjson
import bson
cbson = bson.BSON

assert bson._use_c

import msgpack
import cPickle as pickle
import marshal


import sys
import time
from glob import glob

def fread(path):
    with open(path) as f:
        data = f.read()
    return data

def stdout(string):
    sys.stdout.write(string)
    sys.stdout.flush()

jsonfiles = list(sorted(glob("json/*")))
msgpackfiles = list(sorted(glob("mpk/*")))
marshalfiles = list(sorted(glob("mar/*")))
picklefiles = list(sorted(glob("pickle/*")))
bsonfiles = list(sorted(glob("bson/*")))

# -- preload data
print "preloading files ..."

print "reading json ..."
numfiles = len(jsonfiles)
jsondata = []
for i,file in enumerate(jsonfiles):
    jsondata.append(fread(file))
    stdout('%d of %d\r' % (i+1, numfiles))
stdout("\n")

print "reading msgpack ..."
numfiles = len(msgpackfiles)
msgpackdata = []
for i,file in enumerate(msgpackfiles):
    msgpackdata.append(fread(file))
    stdout('%d of %d\r' % (i+1, numfiles))
stdout("\n")

print "reading marshal ..."
numfiles = len(marshalfiles)
marshaldata = []
for i,file in enumerate(marshalfiles):
    marshaldata.append(fread(file))
    stdout('%d of %d\r' % (i+1, numfiles))
stdout("\n")

print "reading pickle ..."
numfiles = len(picklefiles)
pickledata = []
for i,file in enumerate(picklefiles):
    pickledata.append(fread(file))
    stdout('%d of %d\r' % (i+1, numfiles))
stdout("\n")

print "reading bson ..."
numfiles = len(bsonfiles)
bsondata = []
for i,file in enumerate(bsonfiles):
    bsondata.append(fread(file))
    stdout('%d of %d\r' % (i+1, numfiles))
stdout("\n")

print "\ndecoding test:\n"
td = {}

t0 = time.time()
[json.loads(s) for s in jsondata]
td['json'] = time.time() - t0

t0 = time.time()
[sjson.loads(s) for s in jsondata]
td['simplejson'] = time.time() - t0

t0 = time.time()
[cjson.decode(s) for s in jsondata]
td['cjson'] = time.time() - t0

t0 = time.time()
[ujson.loads(s) for s in jsondata]
td['ujson'] = time.time() - t0

t0 = time.time()
[msgpack.loads(s) for s in msgpackdata]
td['msgpack'] = time.time() - t0

t0 = time.time()
[cbson(s).decode() for s in bsondata]
td['bson'] = time.time() - t0

t0 = time.time()
[marshal.loads(s) for s in marshaldata]
td['marshal'] = time.time() - t0

t0 = time.time()
[pickle.loads(s) for s in pickledata]
td['pickle'] = time.time() - t0

for name in sorted(td):
    print name.ljust(15), "%0.2fs" % td[name]

encoded = [ujson.loads(s) for s in jsondata]

del jsondata
del msgpackdata
del marshaldata
del pickledata
del bsondata


print "\nencoding test:\n"

td = {}

t0 = time.time()
for s in encoded:
    json.dumps(s)
td['json'] = time.time() - t0

t0 = time.time()
for s in encoded:
    sjson.dumps(s)
td['simplejson'] = time.time() - t0

t0 = time.time()
for s in encoded:
    cjson.encode(s)
td['cjson'] = time.time() - t0

t0 = time.time()
for s in encoded:
    ujson.dumps(s)
td['ujson'] = time.time() - t0

t0 = time.time()
for s in encoded:
    msgpack.dumps(s)
td['msgpack'] = time.time() - t0

t0 = time.time()
for s in encoded:
    marshal.dumps(s)
td['marshal'] = time.time() - t0

t0 = time.time()
for s in encoded:
    pickle.dumps(s)
td['pickle'] = time.time() - t0

t0 = time.time()
for s in encoded:
    cbson.encode(s)
td['bson'] = time.time() - t0

for name in sorted(td):
    print name.ljust(15), "%0.2fs" % td[name]