def address_spec(self, stream):
'''
Extract a single address spec from a stream of input, always
operates in strict mode.
'''
# sanity check
if stream is None:
raise ParserException('No input provided to parser.')
if isinstance(stream, str) and not is_pure_ascii(stream):
raise ParserException('ASCII string contains non-ASCII chars.')
# to avoid spinning here forever, limit mailbox length
if len(stream) > MAX_ADDRESS_LENGTH:
raise ParserException('Stream length exceeds maximum allowable ' + \
'address length of ' + str(MAX_ADDRESS_LENGTH) + '.')
self.stream = TokenStream(stream)
addr = self._addr_spec()
if addr:
# optional whitespace
self._whitespace()
# if we hit the end of the stream, we have a valid inbox
if self.stream.end_of_stream():
return addr
return None
def _validate_primary(localpart):
# length check
l = len(localpart)
if l < 4 or l > 32:
return False
# no more than one dot (.)
if localpart.count('.') > 1:
return False
# Grammar: local-part -> alpha { [ dot | underscore ] ( alpha | num ) }"
stream = TokenStream(localpart)
# local-part must being with alpha
alpa = stream.get_token(ALPHA)
if alpa is None:
return False
while True:
# optional dot or underscore token
stream.get_token(DOT) or stream.get_token(UNDERSCORE)
# alpha or numeric
alpanum = stream.get_token(ALPHA) or stream.get_token(NUMERIC)
if alpanum is None:
break
# alpha or numeric must be end of stream
if not stream.end_of_stream():
return False
return True
def _validate_primary(localpart):
# length check
l = len(localpart)
if l < 4 or l > 32:
return False
# no more than one dot (.)
if localpart.count(b'.') > 1:
return False
# Grammar: local-part -> alpha { [ dot | underscore ] ( alpha | num ) }"
stream = TokenStream(localpart)
# local-part must being with alpha
alpa = stream.get_token(ALPHA)
if alpa is None:
return False
while True:
# optional dot or underscore token
stream.get_token(DOT) or stream.get_token(UNDERSCORE)
# alpha or numeric
alpanum = stream.get_token(ALPHA) or stream.get_token(NUMERIC)
if alpanum is None:
break
# alpha or numeric must be end of stream
if not stream.end_of_stream():
return False
return True
def _validate(localpart):
stream = TokenStream(localpart)
# get the gmail base (alpha, num, or dot)
mpart = stream.get_token(GMAIL_BASE)
if mpart is None:
return False
# optional tags
tgs = _tags(stream)
if not stream.end_of_stream():
return False
return True
def _validate(localpart):
stream = TokenStream(localpart)
# get the google base
mpart = stream.get_token(GOOGLE_BASE)
if mpart is None:
return False
# optional tags
tgs = _tags(stream)
if not stream.end_of_stream():
return False
return True
def _validate(localpart):
stream = TokenStream(localpart)
# get the gmail base (alpha, num, or dot)
mpart = stream.get_token(GMAIL_BASE)
if mpart is None:
return False
# optional tags
tgs = _tags(stream)
if not stream.end_of_stream():
return False
return True
def _validate(localpart):
stream = TokenStream(localpart)
# get the hotmail base
mpart = stream.get_token(HOTMAIL_BASE)
if mpart is None:
return False
# optional tags
tgs = _tags(stream)
if not stream.end_of_stream():
return False
return True
def address_spec(self, stream):
'''
Extract a single address spec from a stream of input, always
operates in strict mode.
'''
# sanity check
if stream is None:
raise ParserException('No input provided to parser.')
if isinstance(stream, str) and not is_pure_ascii(stream):
raise ParserException('ASCII string contains non-ASCII chars.')
# to avoid spinning here forever, limit mailbox length
if len(stream) > MAX_ADDRESS_LENGTH:
raise ParserException('Stream length exceeds maximum allowable ' + \
'address length of ' + str(MAX_ADDRESS_LENGTH) + '.')
self.stream = TokenStream(stream)
addr = self._addr_spec()
if addr:
# optional whitespace
self._whitespace()
# if we hit the end of the stream, we have a valid inbox
if self.stream.end_of_stream():
return addr
return None
def _validate_disposable(email_addr):
# Setup for handling EmailAddress type instead of literal string
localpart = email_addr.mailbox
managed = managed_email(email_addr.hostname)
# length check (base + hyphen + keyword)
l = len(localpart)
if l < 3 or l > 65:
return False
# single hyphen
if localpart.count('-') != 1:
return False
# base and keyword length limit
parts = localpart.split('-')
for part in parts:
l = len(part)
if l < 1 or l > 32:
return False
# Grammar: local-part -> alpha { [ alpha | num | underscore ] } hyphen { [ alpha | num ] }
stream = TokenStream(localpart)
# must being with alpha
begin = stream.get_token(ALPHA)
if begin is None and managed:
return False
while True:
# alpha, num, underscore
base = stream.get_token(ALPHANUM) or stream.get_token(UNDERSCORE)
if base is None:
break
# hyphen
hyphen = stream.get_token(HYPHEN)
if hyphen is None:
return False
# keyword must be alpha, num
stream.get_token(ALPHANUM)
if not stream.end_of_stream():
return False
return True
def _validate(localpart):
stream = TokenStream(localpart)
while True:
# get alphanumeric portion
mpart = stream.get_token(ALPHANUM)
if mpart is None:
return False
# get optional dot, must be followed by more alphanumerics
mpart = stream.get_token(DOT)
if mpart is None:
break
# optional tags
tgs = _tags(stream)
if not stream.end_of_stream():
return False
return True
def _validate(localpart):
stream = TokenStream(localpart)
while True:
# get alphanumeric portion
mpart = stream.get_token(ALPHANUM)
if mpart is None:
return False
# get optional dot, must be followed by more alphanumerics
mpart = stream.get_token(DOT)
if mpart is None:
break
# optional tags
tgs = _tags(stream)
if not stream.end_of_stream():
return False
return True
def address_list(self, stream):
'''
Extract a mailbox and/or url list from a stream of input, operates in
strict and relaxed modes.
'''
# sanity check
if not stream:
raise ParserException('No input provided to parser.')
if isinstance(stream, str) and not is_pure_ascii(stream):
raise ParserException('ASCII string contains non-ASCII chars.')
# to avoid spinning here forever, limit address list length
if len(stream) > MAX_ADDRESS_LIST_LENGTH:
raise ParserException('Stream length exceeds maximum allowable ' + \
'address list length of ' + str(MAX_ADDRESS_LIST_LENGTH) + '.')
# set stream
self.stream = TokenStream(stream)
if self.strict is True:
return self._address_list_strict()
return self._address_list_relaxed()
def _validate_disposable(localpart):
# length check (base + hyphen + keyword)
l = len(localpart)
if l < 3 or l > 65:
return False
# single hyphen
if localpart.count('-') != 1:
return False
# base and keyword length limit
parts = localpart.split('-')
for part in parts:
l = len(part)
if l < 1 or l > 32:
return False
# Grammar: local-part -> alpha { [ alpha | num | underscore ] } hyphen { [ alpha | num ] }
stream = TokenStream(localpart)
# must being with alpha
begin = stream.get_token(ALPHA)
if begin is None:
return False
while True:
# alpha, num, underscore
base = stream.get_token(ALPHANUM) or stream.get_token(UNDERSCORE)
if base is None:
break
# hyphen
hyphen = stream.get_token(HYPHEN)
if hyphen is None:
return False
# keyword must be alpha, num
stream.get_token(ALPHANUM)
if not stream.end_of_stream():
return False
return True
def _validate_disposable(localpart):
# length check (base + hyphen + keyword)
l = len(localpart)
if l < 3 or l > 65:
return False
# single hyphen
if localpart.count(b'-') != 1:
return False
# base and keyword length limit
parts = localpart.split(b'-')
for part in parts:
l = len(part)
if l < 1 or l > 32:
return False
# Grammar: local-part -> alpha { [ alpha | num | underscore ] } hyphen { [ alpha | num ] }
stream = TokenStream(localpart)
# must being with alpha
begin = stream.get_token(ALPHA)
if begin is None:
return False
while True:
# alpha, num, underscore
base = stream.get_token(ALPHANUM) or stream.get_token(UNDERSCORE)
if base is None:
break
# hyphen
hyphen = stream.get_token(HYPHEN)
if hyphen is None:
return False
# keyword must be alpha, num
stream.get_token(ALPHANUM)
if not stream.end_of_stream():
return False
return True
def _validate(localpart):
"Grammar: local-part -> alpha { [ dot | underscore ] ( alpha | num ) }"
stream = TokenStream(localpart)
# local-part must being with alpha
alpa = stream.get_token(ALPHA)
if alpa is None:
return False
while True:
# optional dot or underscore token
stream.get_token(DOT) or stream.get_token(UNDERSCORE)
# alpha or numeric
alpanum = stream.get_token(ALPHA) or stream.get_token(NUMERIC)
if alpanum is None:
break
# alpha or numeric must be end of stream
if not stream.end_of_stream():
return False
return True
def _validate(localpart):
stream = TokenStream(localpart)
# localpart must start with alpha
alpa = stream.get_token(ICLOUD_PREFIX)
if alpa is None:
return False
while True:
# optional dot or underscore
stream.get_token(DOT) or stream.get_token(UNDERSCORE)
base = stream.get_token(ICLOUD_BASE)
if base is None:
break
if not stream.end_of_stream():
return False
return True
def _validate_primary(email_addr):
# Setup for handling EmailAddress type instead of literal string
localpart = email_addr.mailbox
managed = managed_email(email_addr.hostname)
# length check
l = len(localpart)
if l < 4 or l > 32:
return False
# no more than one dot (.)
if localpart.count('.') > 1:
return False
# Grammar: local-part -> alpha { [ dot | underscore ] ( alpha | num ) }"
stream = TokenStream(localpart)
# local-part must being with alpha
alpa = stream.get_token(ALPHA)
if alpa is None and managed:
return False
while True:
# optional dot or underscore token
stream.get_token(DOT) or stream.get_token(UNDERSCORE)
# alpha or numeric
alpanum = stream.get_token(ALPHA) or stream.get_token(NUMERIC)
if alpanum is None:
break
# alpha or numeric must be end of stream
if not stream.end_of_stream():
return False
return True
def _validate(localpart):
"Grammar: local-part -> alpha { [ dot | underscore ] ( alpha | num ) }"
stream = TokenStream(localpart)
# local-part must being with alpha
alpa = stream.get_token(ALPHA)
if alpa is None:
return False
while True:
# optional dot or underscore token
stream.get_token(DOT) or stream.get_token(UNDERSCORE)
# alpha or numeric
alpanum = stream.get_token(ALPHA) or stream.get_token(NUMERIC)
if alpanum is None:
break
# alpha or numeric must be end of stream
if not stream.end_of_stream():
return False
return True
def address_list(self, stream):
'''
Extract a mailbox and/or url list from a stream of input, operates in
strict and relaxed modes.
'''
# sanity check
if not stream:
raise ParserException('No input provided to parser.')
if isinstance(stream, str) and not is_pure_ascii(stream):
raise ParserException('ASCII string contains non-ASCII chars.')
# to avoid spinning here forever, limit address list length
if len(stream) > MAX_ADDRESS_LIST_LENGTH:
raise ParserException('Stream length exceeds maximum allowable ' + \
'address list length of ' + str(MAX_ADDRESS_LIST_LENGTH) + '.')
# set stream
self.stream = TokenStream(stream)
if self.strict is True:
return self._address_list_strict()
return self._address_list_relaxed()
def _validate(localpart):
stream = TokenStream(localpart)
# localpart must start with alpha
alpa = stream.get_token(ICLOUD_PREFIX)
if alpa is None:
return False
while True:
# optional dot or underscore
stream.get_token(DOT) or stream.get_token(UNDERSCORE)
base = stream.get_token(ICLOUD_BASE)
if base is None:
break
if not stream.end_of_stream():
return False
return True
class _AddressParser(object):
'''
Do not use _AddressParser directly because it heavily relies on other
private classes and methods and it's interface is not guarenteed, it
will change in the future and possibly break your application.
Instead use the parse() and parse_list() functions in the address.py
module which will always return a scalar or iterable respectively.
'''
def __init__(self, strict=False):
self.stream = None
self.strict = strict
def address_list(self, stream):
'''
Extract a mailbox and/or url list from a stream of input, operates in
strict and relaxed modes.
'''
# sanity check
if not stream:
raise ParserException('No input provided to parser.')
if isinstance(stream, str) and not is_pure_ascii(stream):
raise ParserException('ASCII string contains non-ASCII chars.')
# to avoid spinning here forever, limit address list length
if len(stream) > MAX_ADDRESS_LIST_LENGTH:
raise ParserException('Stream length exceeds maximum allowable ' + \
'address list length of ' + str(MAX_ADDRESS_LIST_LENGTH) + '.')
# set stream
self.stream = TokenStream(stream)
if self.strict is True:
return self._address_list_strict()
return self._address_list_relaxed()
def address(self, stream):
'''
Extract a single address or url from a stream of input, always
operates in strict mode.
'''
# sanity check
if not stream:
raise ParserException('No input provided to parser.')
if isinstance(stream, str) and not is_pure_ascii(stream):
raise ParserException('ASCII string contains non-ASCII chars.')
# to avoid spinning here forever, limit mailbox length
if len(stream) > MAX_ADDRESS_LENGTH:
raise ParserException('Stream length exceeds maximum allowable ' + \
'address length of ' + str(MAX_ADDRESS_LENGTH) + '.')
self.stream = TokenStream(stream)
addr = self._address()
if addr:
# optional whitespace
self._whitespace()
# if we hit the end of the stream, we have a valid inbox
if self.stream.end_of_stream():
return addr
return None
def address_spec(self, stream):
'''
Extract a single address spec from a stream of input, always
operates in strict mode.
'''
# sanity check
if stream is None:
raise ParserException('No input provided to parser.')
if isinstance(stream, str) and not is_pure_ascii(stream):
raise ParserException('ASCII string contains non-ASCII chars.')
# to avoid spinning here forever, limit mailbox length
if len(stream) > MAX_ADDRESS_LENGTH:
raise ParserException('Stream length exceeds maximum allowable ' + \
'address length of ' + str(MAX_ADDRESS_LENGTH) + '.')
self.stream = TokenStream(stream)
addr = self._addr_spec()
if addr:
# optional whitespace
self._whitespace()
# if we hit the end of the stream, we have a valid inbox
if self.stream.end_of_stream():
return addr
return None
def _mailbox_post_processing_checks(self, address):
"Additional post processing checks to ensure mailbox is valid."
parts = address.split('@')
# check if local part is less than 128 octets, the actual
# limit is 64 octets but we double the size here because
# unsubscribe links are frequently longer
lpart = parts[0]
if len(lpart) > 128:
return False
# check if the domain is less than 255 octets
domn = parts[1]
if len(domn) > 253:
return False
# number of labels can not be over 127
labels = domn.split('.')
if len(labels) > 127:
return False
for label in labels:
# check the domain doesn't start or end with - and
# the length of each label is no more than 63 octets
if BAD_DOMAIN.search(label) or len(label) > 63:
return False
return True
def _address_list_relaxed(self):
"Grammar: address-list-relaxed -> address { delimiter address }"
#addrs = []
addrs = flanker.addresslib.address.AddressList()
unparsable = []
# address
addr = self._address()
if addr is None:
# synchronize to the next delimiter (or end of line)
# append the skipped over text to the unparsable list
skip = self.stream.synchronize()
if skip:
unparsable.append(skip)
# if no mailbox and end of stream, we were unable
# return the unparsable stream
if self.stream.end_of_stream():
return [], unparsable
else:
# if we found a delimiter or end of stream, we have a
# valid mailbox, add it
if self.stream.peek(DELIMITER) or self.stream.end_of_stream():
addrs.append(addr)
else:
# otherwise snychornize and add it the unparsable array
skip = self.stream.synchronize()
if skip:
pre = self.stream.stream[:self.stream.stream.index(skip)]
unparsable.append(pre + skip)
# if we hit the end of the stream, return the results
if self.stream.end_of_stream():
return [], [self.stream.stream]
while True:
# delimiter
dlm = self.stream.get_token(DELIMITER)
if dlm is None:
skip = self.stream.synchronize()
if skip:
unparsable.append(skip)
if self.stream.end_of_stream():
break
# address
start_pos = self.stream.position
addr = self._address()
if addr is None:
skip = self.stream.synchronize()
if skip:
unparsable.append(skip)
if self.stream.end_of_stream():
break
else:
# if we found a delimiter or end of stream, we have a
# valid mailbox, add it
if self.stream.peek(DELIMITER) or self.stream.end_of_stream():
addrs.append(addr)
else:
# otherwise snychornize and add it the unparsable array
skip = self.stream.synchronize()
if skip:
sskip = self.stream.stream[start_pos:self.stream.position]
unparsable.append(sskip)
# if we hit the end of the stream, return the results
if self.stream.end_of_stream():
return addrs, unparsable
return addrs, unparsable
def _address_list_strict(self):
"Grammar: address-list-strict -> address { delimiter address }"
#addrs = []
addrs = flanker.addresslib.address.AddressList()
# address
addr = self._address()
if addr is None:
return addrs
if self.stream.peek(DELIMITER):
addrs.append(addr)
while True:
# delimiter
dlm = self.stream.get_token(DELIMITER)
if dlm is None:
break
# address
addr = self._address()
if addr is None:
break
addrs.append(addr)
return addrs
def _address(self):
"Grammar: address -> name-addr-rfc | name-addr-lax | addr-spec | url"
start_pos = self.stream.position
addr = self._name_addr_rfc() or self._name_addr_lax() or \
self._addr_spec() or self._url()
# if email address, check that it passes post processing checks
if addr and isinstance(addr, flanker.addresslib.address.EmailAddress):
if self._mailbox_post_processing_checks(addr.address) is False:
# roll back
self.stream.position = start_pos
return None
return addr
def _url(self):
"Grammar: url -> url"
earl = self.stream.get_token(URL)
if earl is None:
return None
return flanker.addresslib.address.UrlAddress(to_utf8(earl))
def _name_addr_rfc(self):
"Grammar: name-addr-rfc -> [ display-name-rfc ] angle-addr-rfc"
start_pos = self.stream.position
# optional displayname
dname = self._display_name_rfc()
aaddr = self._angle_addr_rfc()
if aaddr is None:
# roll back
self.stream.position = start_pos
return None
if dname:
return flanker.addresslib.address.EmailAddress(dname, aaddr)
return flanker.addresslib.address.EmailAddress(None, aaddr)
def _display_name_rfc(self):
"Grammar: display-name-rfc -> [ whitespace ] word { whitespace word }"
wrds = []
# optional whitespace
self._whitespace()
# word
wrd = self._word()
if wrd is None:
return None
wrds.append(wrd)
while True:
# whitespace
wtsp = self._whitespace()
if wtsp is None:
break
wrds.append(wtsp)
# word
wrd = self._word()
if wrd is None:
break
wrds.append(wrd)
return cleanup_display_name(''.join(wrds))
def _angle_addr_rfc(self):
'''
Grammar: angle-addr-rfc -> [ whitespace ] < addr-spec > [ whitespace ]"
'''
start_pos = self.stream.position
# optional whitespace
self._whitespace()
# left angle bracket
lbr = self.stream.get_token(LBRACKET)
if lbr is None:
# rollback
self.stream.position = start_pos
return None
# addr-spec
aspec = self._addr_spec(True)
if aspec is None:
# rollback
self.stream.position = start_pos
return None
# right angle bracket
rbr = self.stream.get_token(RBRACKET)
if rbr is None:
# rollback
self.stream.position = start_pos
return None
# optional whitespace
self._whitespace()
return aspec
def _name_addr_lax(self):
"Grammar: name-addr-lax -> [ display-name-lax ] angle-addr-lax"
start_pos = self.stream.position
# optional displayname
dname = self._display_name_lax()
aaddr = self._angle_addr_lax()
if aaddr is None:
# roll back
self.stream.position = start_pos
return None
if dname:
return flanker.addresslib.address.EmailAddress(dname, aaddr)
return flanker.addresslib.address.EmailAddress(None, aaddr)
def _display_name_lax(self):
'''
Grammar: display-name-lax ->
[ whitespace ] word { whitespace word } whitespace"
'''
start_pos = self.stream.position
wrds = []
# optional whitespace
self._whitespace()
# word
wrd = self._word()
if wrd is None:
# roll back
self.stream.position = start_pos
return None
wrds.append(wrd)
# peek to see if we have a whitespace,
# if we don't, we have a invalid display-name
if self.stream.peek(WHITESPACE) is None or \
self.stream.peek(UNI_WHITE) is None:
self.stream.position = start_pos
return None
while True:
# whitespace
wtsp = self._whitespace()
if wtsp:
wrds.append(wtsp)
# if we need to roll back the next word
start_pos = self.stream.position
# word
wrd = self._word()
if wrd is None:
self.stream.position = start_pos
break
wrds.append(wrd)
# peek to see if we have a whitespace
# if we don't pop off the last word break
if self.stream.peek(WHITESPACE) is None or \
self.stream.peek(UNI_WHITE) is None:
# roll back last word
self.stream.position = start_pos
wrds.pop()
break
return cleanup_display_name(''.join(wrds))
def _angle_addr_lax(self):
"Grammar: angle-addr-lax -> addr-spec [ whitespace ]"
start_pos = self.stream.position
# addr-spec
aspec = self._addr_spec(True)
if aspec is None:
# rollback
self.stream.position = start_pos
return None
# optional whitespace
self._whitespace()
return aspec
def _addr_spec(self, as_string=False):
'''
Grammar: addr-spec -> [ whitespace ] local-part @ domain [ whitespace ]
'''
start_pos = self.stream.position
# optional whitespace
self._whitespace()
lpart = self._local_part()
if lpart is None:
# rollback
self.stream.position = start_pos
return None
asym = self.stream.get_token(AT_SYMBOL)
if asym is None:
# rollback
self.stream.position = start_pos
return None
domn = self._domain()
if domn is None:
# rollback
self.stream.position = start_pos
return None
# optional whitespace
self._whitespace()
aspec = cleanup_email(''.join([lpart, asym, domn]))
if as_string:
return aspec
return flanker.addresslib.address.EmailAddress(None, aspec)
def _local_part(self):
"Grammar: local-part -> dot-atom | quoted-string"
return self.stream.get_token(DOT_ATOM) or \
self.stream.get_token(QSTRING)
def _domain(self):
"Grammar: domain -> dot-atom"
return self.stream.get_token(DOT_ATOM)
def _word(self):
"Grammar: word -> word-ascii | word-unicode"
start_pos = self.stream.position
# ascii word
ascii_wrd = self._word_ascii()
if ascii_wrd and not self.stream.peek(UNI_ATOM):
return ascii_wrd
# didn't get an ascii word, rollback to try again
self.stream.position = start_pos
# unicode word
return self._word_unicode()
def _word_ascii(self):
"Grammar: word-ascii -> atom | qstring"
wrd = self.stream.get_token(RELAX_ATOM) or self.stream.get_token(QSTRING)
if wrd and not contains_control_chars(wrd):
return wrd
return None
def _word_unicode(self):
"Grammar: word-unicode -> unicode-atom | unicode-qstring"
start_pos = self.stream.position
# unicode atom
uwrd = self.stream.get_token(UNI_ATOM)
if uwrd and isinstance(uwrd, unicode) and not contains_control_chars(uwrd):
return uwrd
# unicode qstr
uwrd = self.stream.get_token(UNI_QSTR, 'qstr')
if uwrd and isinstance(uwrd, unicode) and not contains_control_chars(uwrd):
return u'"{}"'.format(encode_string(None, uwrd))
# rollback
self.stream.position = start_pos
return None
def _whitespace(self):
"Grammar: whitespace -> whitespace-ascii | whitespace-unicode"
return self._whitespace_ascii() or self._whitespace_unicode()
def _whitespace_ascii(self):
"Grammar: whitespace-ascii -> whitespace-ascii"
return self.stream.get_token(WHITESPACE)
def _whitespace_unicode(self):
"Grammar: whitespace-unicode -> whitespace-unicode"
uwhite = self.stream.get_token(UNI_WHITE)
if uwhite and not is_pure_ascii(uwhite):
return uwhite
return None
class _AddressParser(object):
'''
Do not use _AddressParser directly because it heavily relies on other
private classes and methods and it's interface is not guarenteed, it
will change in the future and possibly break your application.
Instead use the parse() and parse_list() functions in the address.py
module which will always return a scalar or iterable respectively.
'''
def __init__(self, strict=False):
self.stream = None
self.strict = strict
def address_list(self, stream):
'''
Extract a mailbox and/or url list from a stream of input, operates in
strict and relaxed modes.
'''
# sanity check
if not stream:
raise ParserException('No input provided to parser.')
if isinstance(stream, str) and not is_pure_ascii(stream):
raise ParserException('ASCII string contains non-ASCII chars.')
# to avoid spinning here forever, limit address list length
if len(stream) > MAX_ADDRESS_LIST_LENGTH:
raise ParserException('Stream length exceeds maximum allowable ' + \
'address list length of ' + str(MAX_ADDRESS_LIST_LENGTH) + '.')
# set stream
self.stream = TokenStream(stream)
if self.strict is True:
return self._address_list_strict()
return self._address_list_relaxed()
def address(self, stream):
'''
Extract a single address or url from a stream of input, always
operates in strict mode.
'''
# sanity check
if not stream:
raise ParserException('No input provided to parser.')
if isinstance(stream, str) and not is_pure_ascii(stream):
raise ParserException('ASCII string contains non-ASCII chars.')
# to avoid spinning here forever, limit mailbox length
if len(stream) > MAX_ADDRESS_LENGTH:
raise ParserException('Stream length exceeds maximum allowable ' + \
'address length of ' + str(MAX_ADDRESS_LENGTH) + '.')
self.stream = TokenStream(stream)
addr = self._address()
if addr:
# optional whitespace
self._whitespace()
# if we hit the end of the stream, we have a valid inbox
if self.stream.end_of_stream():
return addr
return None
def address_spec(self, stream):
'''
Extract a single address spec from a stream of input, always
operates in strict mode.
'''
# sanity check
if stream is None:
raise ParserException('No input provided to parser.')
if isinstance(stream, str) and not is_pure_ascii(stream):
raise ParserException('ASCII string contains non-ASCII chars.')
# to avoid spinning here forever, limit mailbox length
if len(stream) > MAX_ADDRESS_LENGTH:
raise ParserException('Stream length exceeds maximum allowable ' + \
'address length of ' + str(MAX_ADDRESS_LENGTH) + '.')
self.stream = TokenStream(stream)
addr = self._addr_spec()
if addr:
# optional whitespace
self._whitespace()
# if we hit the end of the stream, we have a valid inbox
if self.stream.end_of_stream():
return addr
return None
def _mailbox_post_processing_checks(self, address):
"Additional post processing checks to ensure mailbox is valid."
parts = address.split('@')
# check if local part is less than 256 octets, the actual
# limit is 64 octets but we quadruple the size here because
# unsubscribe links are frequently longer
lpart = parts[0]
if len(lpart) > 256:
return False
# check if the domain is less than 255 octets
domn = parts[1]
if len(domn) > 253:
return False
# number of labels can not be over 127
labels = domn.split('.')
if len(labels) > 127:
return False
for label in labels:
# check the domain doesn't start or end with - and
# the length of each label is no more than 63 octets
if BAD_DOMAIN.search(label) or len(label) > 63:
return False
return True
def _address_list_relaxed(self):
"Grammar: address-list-relaxed -> address { delimiter address }"
#addrs = []
addrs = flanker.addresslib.address.AddressList()
unparsable = []
# address
addr = self._address()
if addr is None:
# synchronize to the next delimiter (or end of line)
# append the skipped over text to the unparsable list
skip = self.stream.synchronize()
if skip:
unparsable.append(skip)
# if no mailbox and end of stream, we were unable
# return the unparsable stream
if self.stream.end_of_stream():
return [], unparsable
else:
# if we found a delimiter or end of stream, we have a
# valid mailbox, add it
if self.stream.peek(DELIMITER) or self.stream.end_of_stream():
addrs.append(addr)
else:
# otherwise snychornize and add it the unparsable array
skip = self.stream.synchronize()
if skip:
pre = self.stream.stream[:self.stream.stream.index(skip)]
unparsable.append(pre + skip)
# if we hit the end of the stream, return the results
if self.stream.end_of_stream():
return [], [self.stream.stream]
while True:
# delimiter
dlm = self.stream.get_token(DELIMITER)
if dlm is None:
skip = self.stream.synchronize()
if skip:
unparsable.append(skip)
if self.stream.end_of_stream():
break
# address
start_pos = self.stream.position
addr = self._address()
if addr is None:
skip = self.stream.synchronize()
if skip:
unparsable.append(skip)
if self.stream.end_of_stream():
break
else:
# if we found a delimiter or end of stream, we have a
# valid mailbox, add it
if self.stream.peek(DELIMITER) or self.stream.end_of_stream():
addrs.append(addr)
else:
# otherwise snychornize and add it the unparsable array
skip = self.stream.synchronize()
if skip:
sskip = self.stream.stream[start_pos:self.stream.
position]
unparsable.append(sskip)
# if we hit the end of the stream, return the results
if self.stream.end_of_stream():
return addrs, unparsable
return addrs, unparsable
def _address_list_strict(self):
"Grammar: address-list-strict -> address { delimiter address }"
#addrs = []
addrs = flanker.addresslib.address.AddressList()
# address
addr = self._address()
if addr is None:
return addrs
if self.stream.peek(DELIMITER):
addrs.append(addr)
while True:
# delimiter
dlm = self.stream.get_token(DELIMITER)
if dlm is None:
break
# address
addr = self._address()
if addr is None:
break
addrs.append(addr)
return addrs
def _address(self):
"Grammar: address -> name-addr-rfc | name-addr-lax | addr-spec | url"
start_pos = self.stream.position
addr = self._name_addr_rfc() or self._name_addr_lax() or \
self._addr_spec() or self._url()
# if email address, check that it passes post processing checks
if addr and isinstance(addr, flanker.addresslib.address.EmailAddress):
if self._mailbox_post_processing_checks(addr.address) is False:
# roll back
self.stream.position = start_pos
return None
return addr
def _url(self):
"Grammar: url -> url"
earl = self.stream.get_token(URL)
if earl is None:
return None
return flanker.addresslib.address.UrlAddress(to_utf8(earl))
def _name_addr_rfc(self):
"Grammar: name-addr-rfc -> [ display-name-rfc ] angle-addr-rfc"
start_pos = self.stream.position
# optional displayname
dname = self._display_name_rfc()
aaddr = self._angle_addr_rfc()
if aaddr is None:
# roll back
self.stream.position = start_pos
return None
if dname:
return flanker.addresslib.address.EmailAddress(dname, aaddr)
return flanker.addresslib.address.EmailAddress(None, aaddr)
def _display_name_rfc(self):
"Grammar: display-name-rfc -> [ whitespace ] word { whitespace word }"
wrds = []
# optional whitespace
self._whitespace()
# word
wrd = self._word()
if wrd is None:
return None
wrds.append(wrd)
while True:
# whitespace
wtsp = self._whitespace()
if wtsp is None:
break
wrds.append(wtsp)
# word
wrd = self._word()
if wrd is None:
break
wrds.append(wrd)
return cleanup_display_name(''.join(wrds))
def _angle_addr_rfc(self):
'''
Grammar: angle-addr-rfc -> [ whitespace ] < addr-spec > [ whitespace ]"
'''
start_pos = self.stream.position
# optional whitespace
self._whitespace()
# left angle bracket
lbr = self.stream.get_token(LBRACKET)
if lbr is None:
# rollback
self.stream.position = start_pos
return None
# addr-spec
aspec = self._addr_spec(True)
if aspec is None:
# rollback
self.stream.position = start_pos
return None
# right angle bracket
rbr = self.stream.get_token(RBRACKET)
if rbr is None:
# rollback
self.stream.position = start_pos
return None
# optional whitespace
self._whitespace()
return aspec
def _name_addr_lax(self):
"Grammar: name-addr-lax -> [ display-name-lax ] angle-addr-lax"
start_pos = self.stream.position
# optional displayname
dname = self._display_name_lax()
aaddr = self._angle_addr_lax()
if aaddr is None:
# roll back
self.stream.position = start_pos
return None
if dname:
return flanker.addresslib.address.EmailAddress(dname, aaddr)
return flanker.addresslib.address.EmailAddress(None, aaddr)
def _display_name_lax(self):
'''
Grammar: display-name-lax ->
[ whitespace ] word { whitespace word } whitespace"
'''
start_pos = self.stream.position
wrds = []
# optional whitespace
self._whitespace()
# word
wrd = self._word()
if wrd is None:
# roll back
self.stream.position = start_pos
return None
wrds.append(wrd)
# peek to see if we have a whitespace,
# if we don't, we have a invalid display-name
if self.stream.peek(WHITESPACE) is None or \
self.stream.peek(UNI_WHITE) is None:
self.stream.position = start_pos
return None
while True:
# whitespace
wtsp = self._whitespace()
if wtsp:
wrds.append(wtsp)
# if we need to roll back the next word
start_pos = self.stream.position
# word
wrd = self._word()
if wrd is None:
self.stream.position = start_pos
break
wrds.append(wrd)
# peek to see if we have a whitespace
# if we don't pop off the last word break
if self.stream.peek(WHITESPACE) is None or \
self.stream.peek(UNI_WHITE) is None:
# roll back last word
self.stream.position = start_pos
wrds.pop()
break
return cleanup_display_name(''.join(wrds))
def _angle_addr_lax(self):
"Grammar: angle-addr-lax -> addr-spec [ whitespace ]"
start_pos = self.stream.position
# addr-spec
aspec = self._addr_spec(True)
if aspec is None:
# rollback
self.stream.position = start_pos
return None
# optional whitespace
self._whitespace()
return aspec
def _addr_spec(self, as_string=False):
'''
Grammar: addr-spec -> [ whitespace ] local-part @ domain [ whitespace ]
'''
start_pos = self.stream.position
# optional whitespace
self._whitespace()
lpart = self._local_part()
if lpart is None:
# rollback
self.stream.position = start_pos
return None
asym = self.stream.get_token(AT_SYMBOL)
if asym is None:
# rollback
self.stream.position = start_pos
return None
domn = self._domain()
if domn is None:
# rollback
self.stream.position = start_pos
return None
# optional whitespace
self._whitespace()
aspec = cleanup_email(''.join([lpart, asym, domn]))
if as_string:
return aspec
return flanker.addresslib.address.EmailAddress(None, aspec)
def _local_part(self):
"Grammar: local-part -> dot-atom | quoted-string"
return self.stream.get_token(DOT_ATOM) or \
self.stream.get_token(QSTRING)
def _domain(self):
"Grammar: domain -> dot-atom"
return self.stream.get_token(DOT_ATOM)
def _word(self):
"Grammar: word -> word-ascii | word-unicode"
start_pos = self.stream.position
# ascii word
ascii_wrd = self._word_ascii()
if ascii_wrd and not self.stream.peek(UNI_ATOM):
return ascii_wrd
# didn't get an ascii word, rollback to try again
self.stream.position = start_pos
# unicode word
return self._word_unicode()
def _word_ascii(self):
"Grammar: word-ascii -> atom | qstring"
wrd = self.stream.get_token(RELAX_ATOM) or self.stream.get_token(
QSTRING)
if wrd and not contains_control_chars(wrd):
return wrd
return None
def _word_unicode(self):
"Grammar: word-unicode -> unicode-atom | unicode-qstring"
start_pos = self.stream.position
# unicode atom
uwrd = self.stream.get_token(UNI_ATOM)
if uwrd and isinstance(uwrd,
unicode) and not contains_control_chars(uwrd):
return uwrd
# unicode qstr
uwrd = self.stream.get_token(UNI_QSTR, 'qstr')
if uwrd and isinstance(uwrd,
unicode) and not contains_control_chars(uwrd):
return u'"{0}"'.format(encode_string(None, uwrd))
# rollback
self.stream.position = start_pos
return None
def _whitespace(self):
"Grammar: whitespace -> whitespace-ascii | whitespace-unicode"
return self._whitespace_ascii() or self._whitespace_unicode()
def _whitespace_ascii(self):
"Grammar: whitespace-ascii -> whitespace-ascii"
return self.stream.get_token(WHITESPACE)
def _whitespace_unicode(self):
"Grammar: whitespace-unicode -> whitespace-unicode"
uwhite = self.stream.get_token(UNI_WHITE)
if uwhite and not is_pure_ascii(uwhite):
return uwhite
return None
