File:Wiktionary Teestube Editfrequenz.svg

# python
# -*- coding: utf8 -*-

import urllib2
import json
from datetime import datetime
from datetime import timedelta
import matplotlib.pyplot
import matplotlib.dates

page = 'Wiktionary:Teestube'
totalcount = 12000

remcount = totalcount
rvcontinue = ''
revlist = []
while remcount > 0:
    if remcount > 500:
        querycount = 500
        remcount -= querycount
    else:
        querycount = remcount
        remcount = 0
    url = 'http://de.wiktionary.org/w/api.php?action=query&prop=revisions&format=json&rvprop=timestamp&rvlimit='+str(querycount)+'&rvdir=newer'
    if rvcontinue != '':
        url += '&rvcontinue='+rvcontinue
    url += '&titles='+page
    response = urllib2.urlopen(url)
    rawdata = response.read()
    jsondata = json.loads(rawdata)
    revlist += jsondata['query']['pages'].values()[0]['revisions']
    try:
        rvcontinue = str(jsondata['query-continue']['revisions']['rvcontinue'])
    except:
        print 'Page history is complete.'
        break

weeks = []
freqs = []
week = datetime.strptime(revlist[0]['timestamp'],'%Y-%m-%dT%H:%M:%SZ')
nextweek = week + timedelta(weeks=1)
weeks.append(week)
freqs.append(0)
for rev in revlist:
    if datetime.strptime(rev['timestamp'],'%Y-%m-%dT%H:%M:%SZ') < nextweek:
        freqs[-1] += 1
    else:
        week = nextweek
        nextweek += timedelta(weeks=1)
        weeks.append(week)
        freqs.append(1)

weeksnum = matplotlib.dates.date2num(weeks)
matplotlib.pyplot.plot_date(weeksnum,freqs,'b-')
matplotlib.pyplot.tight_layout()
matplotlib.pyplot.show()