def parse_mimepart_address_header(mimepart, header_name):
# type: (MimePart, str) -> List[List[str]]
# Header parsing is complicated by the fact that:
# (1) You can have multiple occurrences of the same header;
# (2) Phrases or comments can be RFC2047-style encoded words;
# (3) Everything is terrible.
# Here, for each occurrence of the header in question, we first parse
# it into a list of (phrase, addrspec) tuples and then use flanker to
# decode any encoded words.
# You want to do it in that order, because otherwise if you get a header
# like
# From: =?utf-8?Q?FooCorp=2C=20Inc.=? <*****@*****.**>
# you can end up parsing 'FooCorp, Inc. <*****@*****.**> (note lack of
# quoting) into two separate addresses.
# Consult RFC822 Section 6.1 and RFC2047 section 5 for details.
addresses = set() # type: Set[Tuple[str, str]]
for section in mimepart.headers._v.getall(normalize(header_name)):
for phrase, addrspec in email.utils.getaddresses([section]):
if not addrspec and not phrase:
continue
addresses.add((decode(phrase), decode(addrspec)))
# Return a list of lists because it makes it easier to compare an address
# field to one which has been fetched from the db.
return sorted(list(elem) for elem in addresses)
def parse_address_hval(hval):
"""
Given a address collection header value, parse out the name/email pairs into a list of (name, email)
:param string hval: Email address strings
:return set[string, string]
"""
addresses = set()
for phrase, addrspec in rfc822.AddressList(hval).addresslist:
addresses.add((decode(phrase), decode(addrspec)))
return sorted(elem for elem in addresses)
def get(self, key, default=None):
"""
Returns header value (case-insensitive).
"""
v = self._v.get(normalize(key), default)
if v is not None:
return encodedword.decode(v)
return None
def iteritems(self, raw=False):
"""
Returns iterator header,val pairs in the preserved order.
"""
if raw:
return self._v.iteritems()
return iter([(x[0], encodedword.decode(x[1]))
for x in self._v.iteritems()])
def parse_mimepart_address_header(mimepart, header_name):
# Header parsing is complicated by the fact that:
# (1) You can have multiple occurrences of the same header;
# (2) Phrases or comments can be RFC2047-style encoded words;
# (3) Everything is terrible.
# Here, for each occurrence of the header in question, we first parse
# it into a list of (phrase, addrspec) tuples and then use flanker to
# decode any encoded words.
# You want to do it in that order, because otherwise if you get a header
# like
# From: =?utf-8?Q?FooCorp=2C=20Inc.=? <*****@*****.**>
# you can end up parsing 'FooCorp, Inc. <*****@*****.**> (note lack of
# quoting) into two separate addresses.
# Consult RFC822 Section 6.1 and RFC2047 section 5 for details.
addresses = set()
for section in mimepart.headers._v.getall(normalize(header_name)):
for phrase, addrspec in rfc822.AddressList(section).addresslist:
addresses.add((decode(phrase), decode(addrspec)))
# Return a list of lists because it makes it easier to compare an address
# field to one which has been fetched from the db.
return sorted(list(elem) for elem in addresses)
def parse_header_value(name, val):
if not is_pure_ascii(val):
if parametrized.is_parametrized(name, val):
raise DecodingError("Unsupported value in content- header")
return to_unicode(val)
else:
if parametrized.is_parametrized(name, val):
val, params = parametrized.decode(val)
if name == 'Content-Type':
main, sub = parametrized.fix_content_type(val)
return ContentType(main, sub, params)
else:
return WithParams(val, params)
elif "=?" in val:
# may be encoded word
return encodedword.decode(val)
else:
return val
def parse_header_value(name, val):
if not is_pure_ascii(val):
if parametrized.is_parametrized(name, val):
raise DecodingError("Unsupported value in content- header")
return to_unicode(val)
else:
if parametrized.is_parametrized(name, val):
val, params = parametrized.decode(val)
if name == 'Content-Type':
main, sub = parametrized.fix_content_type(val)
return ContentType(main, sub, params)
else:
return WithParams(val, params)
elif "=?" in val:
# may be encoded word
return encodedword.decode(val)
else:
return val
def __getitem__(self, key):
v = self._v.get(normalize(key), None)
if v is not None:
return encodedword.decode(v)
return None
def getall(self, key):
"""
Returns all header values by the given header name (case-insensitive).
"""
v = self._v.getall(normalize(key))
return [encodedword.decode(x) for x in v]
