testextraction.py 文件源码

def nofoutofplacefeatures(url):


#   pdb.set_trace()

    if url[:4]=="http":
        r = requests.get(url)
    else:
        url="http://"+url
        r  = requests.get(url)

    #r = requests.get(url)
    data = r.text
    data2=r.content

    document, errors = tidy_document(data,
      options={'numeric-entities':1})

    #print document
    #print errors
    #print "Number of Elements Out of Place : " + str(len(errors))
    return len(errors)