|
1
|
|
|
#! /usr/bin/env python |
|
2
|
|
|
# -*- coding: utf-8 -*- |
|
3
|
|
|
# Compatible with python3 and python2 (tested with at least 2.4) |
|
4
|
|
|
|
|
5
|
|
|
# flake8: noqa |
|
6
|
|
|
# pylint: skip-file |
|
7
|
|
|
# Originally from https://github.com/LRGH/emlx2eml |
|
8
|
|
|
|
|
9
|
|
|
import sys |
|
10
|
|
|
import os |
|
11
|
|
|
import logging |
|
12
|
|
|
import struct |
|
13
|
|
|
import email |
|
14
|
|
|
import base64 |
|
15
|
|
|
import mimetypes |
|
16
|
|
|
|
|
17
|
|
|
log = logging.getLogger("emlx2eml") |
|
18
|
|
|
console_handler = logging.StreamHandler() |
|
19
|
|
|
console_handler.setFormatter(logging.Formatter("%(levelname)-5s: %(message)s")) |
|
20
|
|
|
log.addHandler(console_handler) |
|
21
|
|
|
log.setLevel(logging.DEBUG) |
|
22
|
|
|
log.setLevel(logging.ERROR) |
|
23
|
|
|
|
|
24
|
|
|
|
|
25
|
|
|
def find_emlx(input): |
|
26
|
|
|
if os.path.islink(input): |
|
27
|
|
|
return [] |
|
28
|
|
|
elif os.path.isdir(input): |
|
29
|
|
|
files = [] |
|
30
|
|
|
for x in os.listdir(input): |
|
31
|
|
|
files += find_emlx(os.path.join(input, x)) |
|
32
|
|
|
return files |
|
33
|
|
|
elif input.endswith(".emlx"): |
|
34
|
|
|
return [input] |
|
35
|
|
|
else: |
|
36
|
|
|
return [] |
|
37
|
|
|
|
|
38
|
|
|
|
|
39
|
|
|
# Some definitions, to enforce compatibility with python2 and python3 |
|
40
|
|
|
newline = struct.pack("B", 10) |
|
41
|
|
|
if sys.version_info[0] == 2: |
|
42
|
|
|
message_from_bytes = email.message_from_string |
|
43
|
|
|
def message_as_bytes(msg): return msg.as_string(unixfrom=True) |
|
44
|
|
|
else: |
|
45
|
|
|
message_from_bytes = email.message_from_bytes |
|
46
|
|
|
def message_as_bytes(msg): return msg.as_bytes(unixfrom=True) |
|
47
|
|
|
|
|
48
|
|
|
|
|
49
|
|
|
def copy_emlx(emlx, out_dir): |
|
50
|
|
|
# Get the numeric id |
|
51
|
|
|
id = get_numeric_id(emlx) |
|
52
|
|
|
|
|
53
|
|
|
# Create output file |
|
54
|
|
|
if not os.path.exists(out_dir): |
|
55
|
|
|
os.mkdir(out_dir) |
|
56
|
|
|
eml = os.path.join(out_dir, id+".eml") |
|
57
|
|
|
log.debug("Extract %s to %s", emlx, eml) |
|
58
|
|
|
if os.path.exists(eml): |
|
59
|
|
|
log.error("%s already exists", eml) |
|
60
|
|
|
return False |
|
61
|
|
|
# Parse the EMLX file |
|
62
|
|
|
msg = parse_emlx(emlx) |
|
63
|
|
|
msg.set_unixfrom("From emlx2eml Thu Apr 19 00:00:00 2012") |
|
64
|
|
|
# TODO: generate relevant values for unixfrom |
|
65
|
|
|
open(eml, "wb").write(message_as_bytes(msg)) |
|
66
|
|
|
|
|
67
|
|
|
|
|
68
|
|
|
def get_numeric_id(filename): |
|
69
|
|
|
id = os.path.basename(filename) |
|
70
|
|
|
assert(id.endswith(".emlx")) |
|
71
|
|
|
id = id[:-5] |
|
72
|
|
|
if id.endswith(".partial"): |
|
73
|
|
|
id = id[:-8] |
|
74
|
|
|
return id |
|
75
|
|
|
|
|
76
|
|
|
|
|
77
|
|
|
def parse_emlx(filename): |
|
78
|
|
|
# Read file |
|
79
|
|
|
content = open(filename, "rb").read() |
|
80
|
|
|
|
|
81
|
|
|
# Extract parts |
|
82
|
|
|
eol = content.find(newline) |
|
83
|
|
|
length = int(content[:eol]) |
|
84
|
|
|
body = content[eol+1:eol+1+length] |
|
85
|
|
|
# TODO: parse the content of 'plist', e.g. using plistlib |
|
86
|
|
|
# plist = content[eol+1+length:] |
|
87
|
|
|
msg = message_from_bytes(body) |
|
88
|
|
|
|
|
89
|
|
|
# Find where attachments may be |
|
90
|
|
|
id = get_numeric_id(filename) |
|
91
|
|
|
attach_dir = os.path.dirname(filename) |
|
92
|
|
|
if attach_dir == "": |
|
93
|
|
|
attach_dir = "." |
|
94
|
|
|
attach_dir += "/../Attachments/" + id |
|
95
|
|
|
|
|
96
|
|
|
# Make complete eml |
|
97
|
|
|
parse_msg(attach_dir, msg, []) |
|
98
|
|
|
return msg |
|
99
|
|
|
|
|
100
|
|
|
|
|
101
|
|
|
def parse_msg(attach_dir, msg, depth): |
|
102
|
|
|
log.debug("%sPART %s %r of type %s", " "*len(depth), |
|
103
|
|
|
".".join([str(_+1) for _ in depth]), msg, msg.get_content_type()) |
|
104
|
|
|
if msg.is_multipart(): |
|
105
|
|
|
for idx, part in enumerate(msg.get_payload()): |
|
106
|
|
|
parse_msg(attach_dir, part, depth+[idx]) |
|
107
|
|
|
include_attachment(attach_dir, part, depth+[idx]) |
|
108
|
|
|
|
|
109
|
|
|
|
|
110
|
|
|
# When the attachment has no explicit filename, Mail.app generates a name |
|
111
|
|
|
# which we want to guess. The base_filename depends on the OS language at |
|
112
|
|
|
# the time the mail was downloaded. The list below is extracted by parsing |
|
113
|
|
|
# /System/Library/PrivateFrameworks/Notes.framework/Versions/A/Resources/*.lproj/MailCore.strings |
|
114
|
|
|
base_filenames = ( |
|
115
|
|
|
u"مرفق البريد", # ar |
|
116
|
|
|
u"Adjunt de Mail", # ca |
|
117
|
|
|
u"Příloha pošty", # cs |
|
118
|
|
|
u"Postbilag", # da |
|
119
|
|
|
u"Mail-Anhang", # de |
|
120
|
|
|
u"Συνημμένο Mail", # el |
|
121
|
|
|
u"Mail Attachment", # en, en_AU, en_GB |
|
122
|
|
|
u"Archivo adjunto al mensaje", # es |
|
123
|
|
|
u"Archivo adjunto a un correo", # es_419 |
|
124
|
|
|
u"Sähköpostiliite", # fi |
|
125
|
|
|
u"Pièce jointe", # fr, fr_CA |
|
126
|
|
|
u"קובץ מצורף לדואר", # he |
|
127
|
|
|
u"मेल अटैचमेंट", # hi |
|
128
|
|
|
u"E-mail privitak", # hr |
|
129
|
|
|
u"Mail melléklet", # hu |
|
130
|
|
|
u"Lampiran Mail", # id |
|
131
|
|
|
u"Allegato di posta elettronica", # it |
|
132
|
|
|
u"メールの添付ファイル", # ja |
|
133
|
|
|
u"Mail 첨부 파일", # ko |
|
134
|
|
|
u"Lampiran Mail", # ms |
|
135
|
|
|
u"Mail-bijlage", # nl |
|
136
|
|
|
u"E-postvedlegg", # no |
|
137
|
|
|
u"Załącznik poczty", # pl |
|
138
|
|
|
u"Anexo de E-mail", # pt |
|
139
|
|
|
u"Anexo de e‑mail", # pt_PT |
|
140
|
|
|
u"Fișier atașat Mail", # ro |
|
141
|
|
|
u"Вложенный файл Почты", # ru |
|
142
|
|
|
u"Mailová príloha", # sk |
|
143
|
|
|
u"Brevbilaga", # sv |
|
144
|
|
|
u"ไฟล์แนบเมล", # th |
|
145
|
|
|
u"Posta İlişiği", # tr |
|
146
|
|
|
u"Поштове прикріплення", # uk |
|
147
|
|
|
u"Tệp đính kèm của Mail", # vi |
|
148
|
|
|
u"邮件附件", # yue_CN, zh_CN, |
|
149
|
|
|
u"郵件附件", # zh_HK, zh_TW |
|
150
|
|
|
) |
|
151
|
|
|
|
|
152
|
|
|
|
|
153
|
|
|
def mimetypes_guess_extension(mime_type): |
|
154
|
|
|
# We don't want to always use mimetypes.guess_extension, |
|
155
|
|
|
# because it does not always return what is generated by Mail.app, |
|
156
|
|
|
# mainly because multiple extensions can be associated to a single |
|
157
|
|
|
# MIME type. |
|
158
|
|
|
# We prefer to use a hardcoded table. |
|
159
|
|
|
try: |
|
160
|
|
|
return { |
|
161
|
|
|
"text/calendar": u".ics", |
|
162
|
|
|
"image/png": u".png", |
|
163
|
|
|
"image/x-png": u"", |
|
164
|
|
|
"image/gif": u".gif", |
|
165
|
|
|
"image/jpeg": u".jpeg", |
|
166
|
|
|
"image/pjpeg": u".jpg", |
|
167
|
|
|
"image/jpg": u".jpg", |
|
168
|
|
|
"message/rfc822": u".eml", |
|
169
|
|
|
}[mime_type] |
|
170
|
|
|
except KeyError: |
|
171
|
|
|
log.error("Unknown file extension for %r, making a guess...", |
|
172
|
|
|
mime_type) |
|
173
|
|
|
return mimetypes.guess_extension(mime_type) |
|
174
|
|
|
|
|
175
|
|
|
|
|
176
|
|
|
def include_attachment(attach_dir, part, depth): |
|
177
|
|
|
if "X-Apple-Content-Length" not in part: |
|
178
|
|
|
return |
|
179
|
|
|
dirpath = attach_dir + "/" + ".".join([str(_+1) for _ in depth]) |
|
180
|
|
|
file = part.get_filename() |
|
181
|
|
|
if file is None: |
|
182
|
|
|
extension = mimetypes_guess_extension(part.get_content_type()) |
|
183
|
|
|
for base in base_filenames: |
|
184
|
|
|
file = base + extension |
|
185
|
|
|
try: |
|
186
|
|
|
data = open(dirpath+"/"+file, "rb").read() |
|
187
|
|
|
break |
|
188
|
|
|
# python2 raises IOError, python3 raises FileNotFoundError |
|
189
|
|
|
except (IOError, FileNotFoundError): |
|
190
|
|
|
continue |
|
191
|
|
|
else: |
|
192
|
|
|
log.error("%s Unnamed attachment of extension %s not found in %s", |
|
193
|
|
|
" "*len(depth), extension, dirpath) |
|
194
|
|
|
return |
|
195
|
|
|
else: |
|
196
|
|
|
try: |
|
197
|
|
|
data = open(dirpath+"/"+file, "rb").read() |
|
198
|
|
|
except FileNotFoundError: |
|
199
|
|
|
log.error("%s Attachment '%s' not found in %s", |
|
200
|
|
|
" "*len(depth), file, dirpath) |
|
201
|
|
|
return |
|
202
|
|
|
log.debug("%s Attachment '%s' found", " "*len(depth), file) |
|
203
|
|
|
cte = part["Content-Transfer-Encoding"] |
|
204
|
|
|
if cte is None: |
|
205
|
|
|
pass |
|
206
|
|
|
elif cte == "base64": |
|
207
|
|
|
data = base64.b64encode(data) |
|
208
|
|
|
data = newline.join([data[i*76:(i+1)*76] |
|
209
|
|
|
for i in range(len(data)//76+1)]) |
|
210
|
|
|
elif cte == "quoted-printable": |
|
211
|
|
|
# The only example I found was not QP-encoded |
|
212
|
|
|
pass |
|
213
|
|
|
elif cte == "8bit": |
|
214
|
|
|
pass |
|
215
|
|
|
else: |
|
216
|
|
|
log.error("Attachment dir is %s", attach_dir) |
|
217
|
|
|
log.error(" File name is %s", file) |
|
218
|
|
|
log.error(" CTE %r", cte) |
|
219
|
|
|
log.error(" CD %r", part["Content-Disposition"]) |
|
220
|
|
|
part.set_payload(data) |
|
221
|
|
|
|
|
222
|
|
|
|
|
223
|
|
|
if __name__ == "__main__": |
|
224
|
|
|
try: |
|
225
|
|
|
input, out_dir = sys.argv[1:] |
|
226
|
|
|
except ValueError: |
|
227
|
|
|
print("Syntax: emlx2eml.py <source> <output_dir>") |
|
228
|
|
|
print(" <source> can be an EMLX file, or a directory that will") |
|
229
|
|
|
print(" be recursively searched for EMLX files.") |
|
230
|
|
|
sys.exit(1) |
|
231
|
|
|
log.debug("Input %s; Output %s", input, out_dir) |
|
232
|
|
|
for emlx in find_emlx(input): |
|
233
|
|
|
copy_emlx(emlx, out_dir) |
|
234
|
|
|
|
AlexanderWillner /
MailboxCleanup
