Python email quoted-printable encoding problem

I am extracting emails from Gmail using the following:

def getMsgs():
 try:
    conn = imaplib.IMAP4_SSL("imap.gmail.com", 993)
  except:
    print 'Failed to connect'
    print 'Is your internet connection working?'
    sys.exit()
  try:
    conn.login(username, password)
  except:
    print 'Failed to login'
    print 'Is the username and password correct?'
    sys.exit()

  conn.select('Inbox')
  # typ, data = conn.search(None, '(UNSEEN SUBJECT "%s")' % subject)
  typ, data = conn.search(None, '(SUBJECT "%s")' % subject)
  for num in data[0].split():
    typ, data = conn.fetch(num, '(RFC822)')
    msg = email.message_from_string(data[0][1])
    yield walkMsg(msg)

def walkMsg(msg):
  for part in msg.walk():
    if part.get_content_type() != "text/plain":
      continue
    return part.get_payload()

However, some emails I get are nigh impossible for me to extract dates (using regex) from as encoding-related chars such as '=', randomly land in the middle of various text fields. Here's an example where it occurs in a date range I want to extract:

Name: KIRSTI Email: [email protected] Phone #: + 999 99995192 Total in party: 4 total, 0 children Arrival/Departure: Oct 9= , 2010 - Oct 13, 2010 - Oct 13, 2010

Is there a way to remove these encoding characters?


Solution 1:

You could/should use the email.parser module to decode mail messages, for example (quick and dirty example!):

from email.parser import FeedParser
f = FeedParser()
f.feed("<insert mail message here, including all headers>")
rootMessage = f.close()

# Now you can access the message and its submessages (if it's multipart)
print rootMessage.is_multipart()

# Or check for errors
print rootMessage.defects

# If it's a multipart message, you can get the first submessage and then its payload
# (i.e. content) like so:
rootMessage.get_payload(0).get_payload(decode=True)

Using the "decode" parameter of Message.get_payload, the module automatically decodes the content, depending on its encoding (e.g. quoted printables as in your question).

Solution 2:

If you are using Python3.6 or later, you can use the email.message.Message.get_content() method to decode the text automatically. This method supersedes get_payload(), though get_payload() is still available.

Say you have a string s containing this email message (based on the examples in the docs):

Subject: Ayons asperges pour le =?utf-8?q?d=C3=A9jeuner?=
From: =?utf-8?q?Pep=C3=A9?= Le Pew <[email protected]>
To: Penelope Pussycat <[email protected]>,
 Fabrette Pussycat <[email protected]>
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0

    Salut!

    Cela ressemble =C3=A0 un excellent recipie[1] d=C3=A9jeuner.

    [1] http://www.yummly.com/recipe/Roasted-Asparagus-Epicurious-203718

    --Pep=C3=A9
   =20

Non-ascii characters in the string have been encoded with the quoted-printable encoding, as specified in the Content-Transfer-Encoding header.

Create an email object:

import email
from email import policy

msg = email.message_from_string(s, policy=policy.default)

Setting the policy is required here; otherwise policy.compat32 is used, which returns a legacy Message instance that doesn't have the get_content method. policy.default will eventually become the default policy, but as of Python3.7 it's still policy.compat32.

The get_content() method handles decoding automatically:

print(msg.get_content())

Salut!

Cela ressemble à un excellent recipie[1] déjeuner.

[1] http://www.yummly.com/recipe/Roasted-Asparagus-Epicurious-203718

--Pepé

If you have a multipart message, get_content() needs to be called on the individual parts, like this:

for part in message.iter_parts():
    print(part.get_content())

Solution 3:

That's known as quoted-printable encoding. You probably want to use something like quopri.decodestring - http://docs.python.org/library/quopri.html