hidemyass-proxy

hidemyass.py

#hidemyarse
import requests
import BeautifulSoup as BS
headers = {'User-agent':
                            'Mozilla/5.0 (Windows NT 10.0; WOW64; rv:40.0) Gecko/20100101 Firefox/40.0'}

def find_inline_words(checkraw_word):
    inline = []
    none = []
    if len(checkraw_word) == 2:
        inline.append(checkraw_word[1])
        none.append(checkraw_word[0])
    else:
        inline = [v for v in checkraw_word[1:len(checkraw_word)-1:2]]
        none = [j for j in checkraw_word[0:len(checkraw_word)-1:2]]
    in_words = []
    none_words = []
    for word in inline:
        in_words.append(word[1:5])
    for nonew in none:
        none_words.append(nonew[1:5])
    return in_words,none_words

def get_ip(spans,inw,outw):
    ip_adress = ""
    if '\n' in spans:
        spans.remove('\n')
    if ' ' in spans:
        spans.remove(' ')
    for span in spans:
        try:
            style = span['style']
        except (KeyError,TypeError):
            style = None
            try:
                classs = span['class']
            except (KeyError,TypeError):
                classs = None
        if style != None:
            is_inline = style.split(':')[1].strip()
            if is_inline == "inline":
                ip_adress = ip_adress + span.text
        elif classs != None:
            if classs in inw or classs not in outw:
                ip_adress = ip_adress + span.text
        else:
            if type(span) == BS.NavigableString:
                ip_adress = ip_adress + span
    return ip_adress

if __name__ == "__main__":
    url = "http://proxylist.hidemyass.com"
    r = requests.get(url,headers=headers)
    soup = BS.BeautifulSoup(r.content)
    ip_list = []
    ipraw = soup.findAll('tr',{'class':'altshade'}) #raw tr list
    for i in range(len(ipraw)):
        checkraw_word= ipraw[i].style.text.split('\n')
        inw,outw = find_inline_words(checkraw_word) #get inline and none words
        inli = ipraw[i].findAll('span')[1].find('style')
        inli.extract() # remove inline and none tags
        spans = ipraw[i].findAll('span')[1].contents
        host = ipraw[i].findAll('td')[2].text
        ip_list.append(get_ip(spans,inw,outw).strip() + ":" +host )
    with open('hidemyarse.txt', 'w') as f:
        f.write("\n".join(ip_list))
        f.close