1. #!/usr/bin/env python
2. # -*- coding: utf-8 -*-
3. import re
4.  
5. archivo = open("mail.txt", "r")
6. resultado = open("destino.txt","a")
7. emails = list(set([linea.strip() for linea in archivo]))
8.  
9. body_regex = re.compile('''
10.     ^(?!\.)                            # name may not begin with a dot
11.     (
12.       [-a-z0-9!\#$%&'*+/=?^_`{|}~]     # all legal characters except dot
13.       |
14.       (?<!\.)\.                        # single dots only
15.     )+
16.     (?<!\.)$                            # name may not end with a dot
17. ''', re.VERBOSE | re.IGNORECASE)
18. domain_regex = re.compile('''
19.     (
20.       localhost
21.       |
22.       (
23.         [a-z0-9]
24.             # [sub]domain begins with alphanumeric
25.         (
26.           [-\w]*                         # alphanumeric, underscore, dot, hyphen
27.           [a-z0-9]                       # ending alphanumeric
28.         )?
29.       \.                               # ending dot
30.       )+
31.       [a-z]{2,}                        # TLD alpha-only
32.    )$
33. ''', re.VERBOSE | re.IGNORECASE)
34.  
35. def is_valid_email(email):
36.     if not isinstance(email, str) or not email or '@' not in email:
37.         return False
38.    
39.     body, domain = email.rsplit('@', 1)
40.  
41.     match_body = body_regex.match(body)
42.     match_domain = domain_regex.match(domain)
43.  
44.     if not match_domain:
45.         # check for Internationalized Domain Names
46.         # see https://docs.python.org/2/library/codecs.html#module-encodings.idna
47.         try:
48.             domain_encoded = domain.encode('idna').decode('ascii')
49.         except UnicodeError:
50.             return False
51.         match_domain = domain_regex.match(domain_encoded)
52.  
53.     return (match_body is not None) and (match_domain is not None)
54.  
55. for email in emails:
56.   emails = list()
57.   if is_valid_email(email)  and not email in emails:
58.     emails.append(email)
59.     resultado.write(email + '\n')
60.  
61. resultado.close()
62. archivo.close()