Update wiki-site.com spider and list

listsofwikis/mediawiki/wiki-site-spider.py

@@ -1,7 +1,7 @@
-#!/usr/bin/env python2
+#!/usr/bin/env python3
 # -*- coding: utf-8 -*-
 
-# Copyright (C) 2014 WikiTeam developers
+# Copyright (C) 2014-2022 WikiTeam developers
 # This program is free software: you can redistribute it and/or modify
 # it under the terms of the GNU General Public License as published by
 # the Free Software Foundation, either version 3 of the License, or
@@ -17,6 +17,7 @@
 
 import re
 import requests
+from urllib import parse
 
 def main():
     headers = {
@@ -27,12 +28,15 @@ def main():
         'http://www.wiki.co.il/active-wiki-all.html',
         'http://www.wiki.co.il/active-wiki-en.html',
     ]
+    wikis = []
     for url in urls:
-        r = requests.get(url, headers=headers)
-        raw = r.text
-        m = re.findall(ur'<td><a href="([^>]+?)"', raw)
-        for i in m:
-            print i
+        req = requests.get(url, headers=headers)
+        wikis.extend(re.findall(r'<td><a href="([^>]+?)"', req.text))
+
+    wikis.sort()
+    with open('wiki-site.com', 'w') as f:
+        for wiki in wikis:
+            f.write(parse.urljoin(wiki, 'api.php') + '\n')
 
 if __name__ == '__main__':
     main()