#!/usr/bin/python
from Harvester import WebParser
url = "http://www.mrhinkydink.com/proxies%s.htm"
regex = r'\<td\>(\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})\</td\>\n*\<td\>(\d{2,5})\</td\>'
pages = (None, 2, 5)
wp = WebParser(url, regex, pages)
for remote in wp.remotes():
wp.scan(remote)
#!/usr/bin/python
from Harvester import WebParser
url = "http://www.xroxy.com/proxylist.php?port=&type=All_http&pnum=%d"
ip_regex = r"View this Proxy details.*\>(\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})"
port_regex = r"Select proxies with port number (\d{2,5})"
pages = 250
wp = WebParser(url, (ip_regex, port_regex), pages)
for remote in wp.remotes():
wp.scan(remote)
#!/usr/bin/python
from Harvester import WebParser
url = "http://free-proxy.cz/en/proxylist/main/%d"
# </div> 117.175.231.117</td>
# <td><span class="fport">8123</span></td>
ip_regex = r"\<\/div\> (\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})\<\/td\>"
port_regex = r'\<span\ class\=\"fport\"\>(\d{2,5})\<\/span\>'
pages = 179
wp = WebParser(url,(ip_regex,port_regex),pages)
for remote in wp.remotes():
wp.scan(remote)