def summaryFromURL(url): # Grab the JSON entry = JSONFromURL(url) # Retrieve the overview entryText = entry["overview"].get_text() # Ensure the overview is well formed overview = BeautifulSoup(entryText) # Summarize and return the overview return getSummary(overview)
def parseWiki(search): url = 'http://en.wikipedia.org/w/api.php?format=json&action=opensearch&search=' + search + '&prop=revisions&rvprop=content' r = urlopen(url).read() dict1 = json.loads(r) dict2 = dict1['1'][0] url2 = 'http://en.wikipedia.org/wiki/' + dict2 ''' #Finding page id: dict3 = dict2.split(" ") if dict3.length() > 1: dict3 = dict2[0] for i in range(1, company.length()): dict3 = dict3 + '_' + dict2.split(" ")[i] url2 = 'http://en.wikipedia.org/wiki/' + dict3 http://en.wikipedia.org/w/api.php?action=query&titles=Albert%20Einstein&prop=info&format=jsonfm ''' #parse article p = 'http://en.wikipedia.org/w/api.php?action=parse&prop=text&page=' + dict2 + '&format=json' p1 = urlopen(p).read() p2 = json.loads(p1) p3 = p2['parse']['text']['*'] p4 = BeautifulSoup(p3) p5 = p4.find_all('p') p6 = p5[0] p7 = p6.getText() #create bitly shortlink shortapi = "https://api-ssl.bitly.com/v3/shorten?access_token=794e02fd047d7fcc0c44543742d0f471e2f9ebc8&longUrl=" + url2 shortLink = json.loads( urlopen(shortapi).read() ) shortened = shortLink['data']['url'] # Return a list of messages to send return separate(getSummary(p7 ) + str(shortened)[7:15]+str(shortened)[15:])