def test_from_regex (self):
# Testing new regex from bug #1633678
f = open(self._path, 'w')
f.write("""From [email protected] Mon May 31 13:24:50 2004 +0200
Subject: message 1
body1
From [email protected] Mon May 31 13:24:50 2004 -0200
Subject: message 2
body2
From [email protected] Mon May 31 13:24:50 2004
Subject: message 3
body3
From [email protected] Mon May 31 13:24:50 2004
Subject: message 4
body4
""")
f.close()
box = mailbox.UnixMailbox(open(self._path, 'r'))
self.assert_(len(list(iter(box))) == 4)
def main():
import mailbox
print('Reading input file...')
f = open("mbox", 'rb')
mbox = mailbox.UnixMailbox(f)
msglist = []
while 1:
msg = mbox.next()
if msg is None:
break
m = make_message(msg)
msglist.append(m)
f.close()
print('Threading...')
subject_table = thread(msglist)
# Output
L = subject_table.items()
L.sort()
for subj, container in L:
print_container(container)
def _selectBox(self):
# mBox Strict
if self.boxtype.get() == self.boxtyps[0]:
self.mb = mailbox.UnixMailbox(file(self.mailbox.get(), 'r'))
self.Disp(self.boxtype.get(), " at location ", self.mailbox.get(),
" Opened Successfully.")
# mBox Loose
elif self.boxtype.get() == self.boxtyps[1]:
self.mb = mailbox.PortableUnixMailbox(file(self.mailbox.get(),
'r'))
self.Disp(self.boxtype.get(), " at location ", self.mailbox.get(),
" Opened Successfully.")
# MailDir
elif self.boxtype.get() == self.boxtyps[2]:
self.mb = mailbox.Maildir(os.path.dirname(self.mailbox.get()))
self.Disp(self.boxtype.get(), " at location ",
os.path.dirname(self.mailbox.get()),
" Opened Successfully.")
# MMDF
elif self.boxtype.get() == self.boxtyps[3]:
self.mb = mailbox.MmdfMailbox(file(self.mailbox.get(), 'r'))
self.Disp(self.boxtype.get(), " at location ", self.mailbox.get(),
" Opened Successfully.")
# MH
elif self.boxtype.get() == self.boxtyps[4]:
self.mb = mailbox.MHMailbox(file(self.mailbox.get(), 'r'))
self.Disp(self.boxtype.get(), " at location ", self.mailbox.get(),
" Opened Successfully.")
# Babyl
elif self.boxtype.get() == self.boxtyps[5]:
self.mb = mailbox.BabylMailbox(file(self.mailbox.get(), 'r'))
self.Disp(self.boxtype.get(), " at location ", self.mailbox.get(),
" Opened Successfully.")
#Unknown File Type
else:
self.Disp("*** I don't know about that file type.")
self.running = 2
def load_from_file():
if tornado.options.options.init:
delete_index()
create_index()
if tornado.options.options.skip:
logging.info("Skipping first %d messages from mbox file" %
tornado.options.options.skip)
count = 0
upload_data = list()
logging.info("Starting import from file %s" %
tornado.options.options.infile)
mbox = mailbox.UnixMailbox(open(tornado.options.options.infile, 'rb'),
email.message_from_file)
emailParser = DelegatingEmailParser(
[AmazonEmailParser(), SteamEmailParser()])
for msg in mbox:
count += 1
if count < tornado.options.options.skip:
continue
item = convert_msg_to_json(msg)
if item:
upload_data.append(item)
if len(upload_data) == tornado.options.options.batch_size:
upload_batch(upload_data)
upload_data = list()
# upload remaining items in `upload_batch`
if upload_data:
upload_batch(upload_data)
logging.info("Import done - total count %d" % count)
def setupUpdatesMethod(self, numUpdates):
""" this method prepares a datastructure for the updates test.
we are reading the first n mails from the primary mailbox.
they are used for the update test
"""
i = 0
dict = {}
mb = mailbox.UnixMailbox(open(mbox, "r"))
msg = mb.next()
while msg and i < numUpdates:
obj = testMessage(msg)
mid = msg.dict.get("message-id", None)
if mid:
dict[mid] = obj
i = i + 1
msg = mb.next()
return dict
#!/usr/bin/env python
import mailbox # read various mailbox formats (incl. Unix format)
import smtplib # tools for sending mail
import string
# open a part of my outgoing mailbox, for resending:
mbox = mailbox.UnixMailbox(open("testbox", "r"))
while 1:
msg = mbox.next() # grab next message
if not msg: break
# can extract mail header fields like a dictionary,
# e.g. msg['To'], msg['From'], msg['cc'], msg['date'] etc.
# msg.keys() lists all keys for this message
# msg.fp.read() gets the body of the message
# str(msg) is the header of the message
# send message to msg['To'], those on the cc list, and myself:
to = [msg['To']]
if msg.has_key('cc'):
to += map(string.strip, msg['cc'].split(','))
to.append(msg['From']) # add myself
message = str(msg) + """
Due to an error with my email connection, the email I sent
you on %s may not have reached you.
A copy of the message is inserted below.
I apologize if you end up with multiple copies of this message.
===============================================================================
json_part = {}
if part.get_content_maintype() == 'multipart':
continue
json_part['contentType'] = part.get_content_type()
content = part.get_payload(decode=False).decode('utf-8', 'ignore')
json_part['content'] = cleanContent(content)
json_msg['parts'].append(json_part)
# Finally, convert date from asctime to milliseconds since epoch using the
# $date descriptor so it imports "natively" as an ISODate object in MongoDB
then = parse(json_msg['Date'])
millis = int(
time.mktime(then.timetuple()) * 1000 + then.microsecond / 1000)
json_msg['Date'] = {'$date': millis}
return json_msg
mbox = mailbox.UnixMailbox(open(MBOX, 'rb'), email.message_from_file)
# Write each message out as a JSON object on a separate line
# for easy import into MongoDB via mongoimport
f = open(OUT_FILE, 'w')
for msg in gen_json_msgs(mbox):
if msg != None:
f.write(json.dumps(msg, cls=Encoder) + '\n')
f.close()
#!/usr/bin/python
# This program removes duplicated messages from a mailbox
import mailbox
fp = open('/var/spool/mail/gregb')
mb = mailbox.UnixMailbox(fp)
outfp = open('/tmp/mailbox-gregb', 'w')
prev = None
msg = mb.next()
while msg is not None:
try:
msgid = msg['Message-ID']
except:
msgid = None
if (prev is not None) and (msgid is not None) and (prev == msgid):
print "Skipping duplicate message", msgid
msg = mb.next()
continue
msg.fp.seek(msg.startofheaders)
outfp.write(msg.fp.read())
print "Wrote", msgid
msg = mb.next()
if __name__ == '__main__':
import networkx as nx
try:
import matplotlib.pyplot as plt
except:
pass
if len(sys.argv)==1:
file="unix_email.mbox"
else:
file=sys.argv[1]
fp=open(file,"r")
mbox = mailbox.UnixMailbox(fp, msgfactory) # parse unix mailbox
G=nx.MultiDiGraph() # create empty graph
# parse each messages and build graph
for msg in mbox: # msg is python email.Message.Message object
(source_name,source_addr) = parseaddr(msg['From']) # sender
# get all recipients
# see http://www.python.org/doc/current/lib/module-email.Utils.html
tos = msg.get_all('to', [])
ccs = msg.get_all('cc', [])
resent_tos = msg.get_all('resent-to', [])
resent_ccs = msg.get_all('resent-cc', [])
all_recipients = getaddresses(tos + ccs + resent_tos + resent_ccs)
# now add the edges for this mail message
for (target_name,target_addr) in all_recipients:
def inc():
import Zope2, thread
min, max = atoi(sys.argv[3]), atoi(sys.argv[4])
count = max - min
try:
threads = atoi(sys.argv[5])
except:
threads = 1
wait = 0
else:
try:
wait = atof(sys.argv[6])
except:
wait = 0.25
wait = wait * 2
count = count / threads
max = min + count
omin = min
db = Zope2.DB
size = db.getSize()
mem = VmSize()
t = time.time()
c = time.clock()
mbox = sys.argv[2]
argss = []
for i in range(threads):
amin = min + i * count
dest = 'maili%s' % amin
initmaili(dest)
f = open(mbox)
mb = mailbox.UnixMailbox(f)
j = 0
while j < amin:
mb.next()
j = j + 1
lock = thread.allocate_lock()
lock.acquire()
def returnf(t, c, size, mem, r, lock=lock):
print c, r
lock.release()
argss.append((lock, (dest, mb, f, count, wait), returnf))
for lock, args, returnf in argss:
thread.start_new_thread(do, (Zope2.DB, loadinc, args, returnf))
for lock, args, returnf in argss:
lock.acquire()
t = time.time() - t
c = time.clock() - c
size = db.getSize() - size
mem = VmSize() - mem
print t, c, size, mem
#hist("%s-%s-%s" % (omin, count, threads))
Zope2.DB.close()
latest = previous
last_processed = None
# process updated mbox files
for file in glob(archive):
if int(previous) >= int(os.stat(file).st_mtime): continue
# open gzipped/raw file
if file.endswith('.gz'):
fh = gzip.open(file)
else:
fh = open(file)
# process each multipart message in the mailbox
for msg in iter(mailbox.UnixMailbox(fh, email.message_from_file)):
last_processed = msg['Date']
if msg.is_multipart():
detach(msg)
elif '919-573-9199' in msg.get_payload():
if '-----BEGIN PGP SIGNATURE-----' in msg.get_payload().split(
"\n"):
msg.add_header('Content-Disposition',
'attachment',
filename='pgp.txt')
wrapper = email.message.Message()
wrapper.attach(msg)
for header in msg.keys():
wrapper[header] = msg[header]
detach(wrapper)
import mailbox
mb = mailbox.UnixMailbox(open("/var/spool/mail/effbot"))
while 1:
msg = next(mb)
if not msg:
break
for k, v in list(msg.items()):
print(k, "=", v)
body = msg.fp.read()
print(len(body), "bytes in body")
## subject = for he's a ...
## message-id = <*****@*****.**>
## received = (from [email protected])
## by spam.egg (8.8.7/8.8.5) id CAA03202
## for effbot; Fri, 15 Oct 1999 02:27:36 +0200
## from = Fredrik Lundh <*****@*****.**>
## date = Fri, 15 Oct 1999 12:35:36 +0200
## to = [email protected]
## 1295 bytes in body
'''
mailbox 模块
mailbox 模块用来处理各种不同类型的邮箱格式
大部分邮箱格式使用文本文件储存纯 RFC 822 信息, 用分割行区别不同的信息.
'''
import mailbox
help(mailbox)
mb = mailbox.UnixMailbox(open(''))
def __init__(self, reader):
sqmail.gui.fetcher.Fetcher.__init__(self, reader, "Spool Read")
filename = sqmail.preferences.get_incomingpath()
self.msg("Using spool file "+filename)
self.msg("Locking spool file")
rv = os.system("lockfile-create --retry 1 "+filename)
if rv:
self.msg("Failed to lock spool file, aborting")
self.do_abort()
return
self.msg("Opening spool file")
fp = open(filename, "r+")
fp.seek(0, 2)
len = fp.tell()
fp.seek(0, 0)
if (len == 0):
self.msg("Spool file empty. Aborting.")
fp.close()
os.system("lockfile-remove "+filename)
self.do_abort()
return
mbox = mailbox.UnixMailbox(fp)
count = 0
self.msg("Reading messages")
while 1:
self.progress(fp.tell(), len)
msg = sqmail.message.Message()
mboxmsg = mbox.next()
if not mboxmsg:
break
msg.loadfrommessage(mboxmsg)
msg.savealltodatabase()
count = count + 1
if self.abort:
self.msg("Aborted!")
self.msg("(Duplicate messages remain in spool file.)")
break
self.msg(str(count)+" message(s) read")
if not self.abort:
if sqmail.preferences.get_deleteremote():
self.msg("All messages read; truncating spool file")
fp.truncate(0)
else:
self.msg("All messages read. Leaving mail in spool file. " \
"(Fetching again will result in duplicate messages "\
"in your database.)")
self.msg("Closing and unlocking spool file")
fp.close()
os.system("lockfile-remove "+filename)
if not self.abort:
self.do_abort()
