jsons2csv.py

gistfile1.txt

# -*- coding: utf-8 -*-

import json, codecs, os

srcdir = 'perdate/'
destdir = 'perdate_tsv/'

if not os.path.exists(destdir):
    os.mkdir(destdir)

for month in range(12,13):
    month = '%02d' % month

    print month
    os.mkdir(destdir + month)
    for date in range(1,31):
        date = '%02d' % date

        print date

        jsonFile = open(srcdir + month + '/' + date + '.jsons', 'r')
        output = open(destdir + month + '/' + date + '.tsv', 'w')

        line = jsonFile.readline()

        while line:
            data = json.loads(line)
            jsonData = data['textPayload'].encode('utf-8')
            output.write(jsonData + "\n")
            line = jsonFile.readline()

        jsonFile.close
        output.close