SourceBot

Description

Le bot détecte l'absence éventuelle d'une source dans une ligne biographique et la signale Il met à jour une page FactChecking qui liste les entrées non sourcées en associant une référence particulière à cet inconnue Il scanne régulièrement la page, si la sourcée est ajoutée il a remet dans la biographie.

Code

1. coding: utf-8

1. In[5]:

import urllib import requests from bs4 import BeautifulSoup import re

user='Vlaedr' passw='Alextall007' baseurl='http://wikipast.epfl.ch/wikipast/' summary='Wikipastbot update'

1. names=['Daniel Brélaz']

1. In[6]:

1. Login request

payload={'action':'query','format':'json','utf8':,'meta':'tokens','type':'login'} r1=requests.post(baseurl + 'api.php', data=payload)

1. login confirm

login_token=r1.json()['query']['tokens']['logintoken'] payload={'action':'login','format':'json','utf8':,'lgname':user,'lgpassword':passw,'lgtoken':login_token} r2=requests.post(baseurl + 'api.php', data=payload, cookies=r1.cookies)

1. get edit token2

params3='?format=json&action=query&meta=tokens&continue=' r3=requests.get(baseurl + 'api.php' + params3, cookies=r2.cookies) edit_token=r3.json()['query']['tokens']['csrftoken']

edit_cookie=r2.cookies.copy() edit_cookie.update(r3.cookies)

1. In[7]:

fullbios = name = 'Biographies' result=requests.post(baseurl+'api.php?action=query&titles='+name+'&export&exportnowrap') soup=BeautifulSoup(result.text, "lxml")

1. soup=BeautifulSoup(result.text)

code= for primitive in soup.findAll("text"):

   code+=primitive.string

fullbios = code

1. In[8]:

allnames = [] pattern = '\* ?\[\[([^\]]*)\]\].*' p = re.compile(pattern) bioNames = fullbios.split('\n') for c in bioNames:

 tk = c.split('\n')
 for t in tk:
   if t:
       match = p.match(t)
       if match:
           allnames.append(match.group(1))
           

1. In[9]:

fullCode = [] for name in allnames:

   result=requests.post(baseurl+'api.php?action=query&titles='+name+'&export&exportnowrap')
   soup=BeautifulSoup(result.text, "lxml")
   #soup=BeautifulSoup(result.text)
   code=
   for primitive in soup.findAll("text"):
       code += primitive.string
   fullCode.append((code, name))

1. print(fullCode)

1. In[23]:

1. the argument is the plain text of the page
2. this fonction returns a tuple:
3. first element is a boolean: true if all entries are sourced
4. false if sources are missing
5. second element is a list of : (modification) all dates where the source is wrong.
6. int: all the bad lines indexes

def findBadPage(pageText):

   # get all the lines
   tokens = []
   tk = pageText.split('\n')
   for t in tk:
       if t:
           if t[0] == '*':
               tokens.append(t)
      
   #check if line is sourced
   r = '\*.*\[https?:\/\/w?w?w?\.?letemps[^\]]*\].*'
   p = re.compile(r)
   dates_isolator_expr = ['\* \[\[(.*)\]\] \/', '\*\[\[(.*)\]\]\:', '\*\[\[(.*)\]\]\/']
   index = 0
   allSourced = True
   wrongDatesSources = []
   
   for t in tokens:
       match = p.match(t)
       if not match:
           print(t)
           allSourced = False
           #d = re.compile(r'(\* \[\[(.*)\]\] \/|\*\[\[(.*)\]\]\/|\[\[(.*)\]\]\:)')
           count = 0;
           didmatch = False
           for i in dates_isolator_expr :
               count +=1
              
               d = re.compile(i)
               match = d.match(t)
               if match:
                   didmatch = True
                   wrongDatesSources.append(match.group(1))
                   #print(match.group(1))
           if count == 3 and not didmatch :
               wrongDatesSources.append('false source at line: ' + str(index))
               otherSource = '\*.*\[(https?:\/\/.*)\].*'
               pOth = re.compile(otherSource)
               match = pOth.match(t)
       index +=1
   return (allSourced, wrongDatesSources)

1. In[24]:

content = '\n' content += 'Cette page liste toutes les biographies ayant des entrées non sourcées.' content += '\n==Bad page==\n'

badpage = [] wrongDatesSources = [] ok =True for (c, name) in fullCode:

   (ok, wrongDatesSources) = findBadPage(c)
   if not ok:
       print(name)
       badpage.append(name)
       content += name
       content += '\n The wrong entries in this page are the sources with the following dates: ' 
       #indexPages = []
       content += str(wrongDatesSources) 
       content += '\n\n'

1. In[22]:

content = '\n' content += 'Cette page liste toutes les biographies ayant des entrées non sourcées.' content += '\n==Bad page==' content += '\n[[' + badpage[0] + ']]' payload = {'action':'edit','assert':'user','format':'json','utf8':,'text':content,

          'summary':summary,'title':'FactChecking','token':edit_token}

r4=requests.post(baseurl+'api.php',data=payload,cookies=edit_cookie)