JGmNS9vk

1# -*- coding:utf8 -*-
2"""
3Author : Myth
4Date   : 18/1/11
5Email  : email4myth at gmail.com
6"""
7
8import email
9import email.header
10import hashlib
11
12import chardet
13
14
15def choose_alternative_part(subparts):
16    return sorted(subparts, key=lambda m: 1 if m.get_content_subtype() == 'html' else 0, reverse=True)[0]
17
18
19def walk(mail):
20    if mail.is_multipart():
21        subparts = mail.get_payload()
22        content_type = mail.get_content_type()
23        if content_type and content_type.lower() == 'multipart/alternative':
24            prefer_subpart = choose_alternative_part(subparts)
25            for subpart in walk(prefer_subpart):
26                yield subpart
27        else:
28            for subpart in subparts:
29                for subsubpart in walk(subpart):
30                    yield subsubpart
31    else:
32        yield mail
33
34
35def decode_header(header):
36    rmsg = ''
37    header = header.replace('"', '')
38    for ret in email.header.decode_header(header):
39        msg, charset = ret
40        msg = decode_str(msg, charset)
41        if not rmsg:
42            rmsg += msg
43        else:
44            rmsg += ' ' + msg
45    return rmsg
46
47
48def is_attachment(msg):
49    return bool(msg.get_filename())
50
51
52def split_letter_and_attach(msgs):
53    letters = []
54    attachments = []
55
56    for msg in msgs:
57        if is_attachment(msg):
58            attachments.append(msg)
59        elif msg.get_content_maintype().lower() == 'text':  # 非附件只取文本类内容
60            letters.append(msg)
61    return letters, attachments
62
63
64def decode_str(s, charset, errors='ignore'):
65    if not s:
66        return s
67
68    if isinstance(s, unicode):
69        return s
70
71    if not charset:
72        ret = chardet.detect(s)
73        # {'confidence': 0.99, 'encoding': 'GB2312'}
74        if ret:
75            charset = ret.get('encoding')
76
77        if not charset:
78            charset = 'utf-8'
79
80    return s.decode(charset, errors=errors)
81
82
83def merge_letters(letters):
84    content = u''
85
86    for letter in letters:
87        payload = letter.get_payload(decode=True)
88        if not payload:
89            continue
90
91        charset = letter.get_content_charset()
92        payload = decode_str(payload, charset)
93
94        content += payload
95        content += '\r\n\r\n'
96
97    return content
98
99
100def content_md5(content):
101    return hashlib.md5(content).hexdigest()
102
103
104def save_attachment(attachment):
105    content = attachment.get_payload(decode=True)
106    filename = decode_header(attachment.get_filename())
107    md5 = content_md5(content)
108    return {
109        'filename': filename,
110        'md5': md5,
111        'content': content
112    }
113
114
115def parse_raw_mail(raw_mail, truncated):
116    try:
117        mail = email.message_from_string(raw_mail.lstrip())
118    except Exception as e:
119        print "error happened when get message from str: %s" % e
120        return {
121            'content': '',
122            'attachments': []
123        }
124    msgs = list(walk(mail))
125    if not msgs:
126        return {
127            'content': '',
128            'attachments': []
129        }
130
131    if truncated and is_attachment(msgs[-1]):  # ignore last truncated attachment
132        msgs.pop()
133
134    letters, attachments = split_letter_and_attach(msgs)
135
136    letter_content = merge_letters(letters)
137
138    attach_infos = []
139    for attach in attachments:
140        attach_info = save_attachment(attach)
141        attach_infos.append(attach_info)
142
143    mail_info = {
144        'content': letter_content,
145        'attachments': attach_infos
146    }
147    return mail_info