def extract_body(mail, types=None):
"""
returns a body text string for given mail.
If types is `None`, 'text/*' is used:
In case mail has a 'text/html' part, it is prefered over
'text/plain' parts.
:param mail: the mail to use
:type mail: :class:`email.Message`
:param types: mime content types to use for body string
:type types: list of str
"""
html = list(typed_subpart_iterator(mail, 'text', 'html'))
# if no specific types are given, we favor text/html over text/plain
drop_plaintext = False
if html and not types:
drop_plaintext = True
body_parts = []
for part in mail.walk():
ctype = part.get_content_type()
if types is not None:
if ctype not in types:
continue
enc = part.get_content_charset() or 'ascii'
raw_payload = part.get_payload(decode=True)
if part.get_content_maintype() == 'text':
raw_payload = string_decode(raw_payload, enc)
if ctype == 'text/plain' and not drop_plaintext:
body_parts.append(string_sanitize(raw_payload))
else:
#get mime handler
handler = get_mime_handler(ctype, key='view',
interactive=False)
if handler:
#open tempfile. Not all handlers accept stuff from stdin
tmpfile = tempfile.NamedTemporaryFile(delete=False,
suffix='.html')
#write payload to tmpfile
if part.get_content_maintype() == 'text':
tmpfile.write(raw_payload.encode('utf8'))
else:
tmpfile.write(raw_payload)
tmpfile.close()
#create and call external command
cmd = handler % tmpfile.name
cmdlist = shlex.split(cmd.encode('utf-8', errors='ignore'))
rendered_payload, errmsg, retval = helper.call_cmd(cmdlist)
#remove tempfile
os.unlink(tmpfile.name)
if rendered_payload: # handler had output
body_parts.append(string_sanitize(rendered_payload))
elif part.get_content_maintype() == 'text':
body_parts.append(string_sanitize(raw_payload))
# else drop
return '\n\n'.join(body_parts)
def decode_header(header, normalize=False):
"""
decode a header value to a unicode string
values are usually a mixture of different substrings
encoded in quoted printable using diffetrent encodings.
This turns it into a single unicode string
:param header: the header value
:type header: str in us-ascii
:param normalize: replace trailing spaces after newlines
:type normalize: bool
:rtype: unicode
"""
valuelist = email.header.decode_header(header)
decoded_list = []
for v, enc in valuelist:
v = string_decode(v, enc)
decoded_list.append(string_sanitize(v))
value = u' '.join(decoded_list)
if normalize:
value = re.sub(r'\n\s+', r' ', value)
return value