jpayne@7: from . import idnadata
jpayne@7: import bisect
jpayne@7: import unicodedata
jpayne@7: import re
jpayne@7: from typing import Union, Optional
jpayne@7: from .intranges import intranges_contain
jpayne@7: 
jpayne@7: _virama_combining_class = 9
jpayne@7: _alabel_prefix = b'xn--'
jpayne@7: _unicode_dots_re = re.compile('[\u002e\u3002\uff0e\uff61]')
jpayne@7: 
jpayne@7: class IDNAError(UnicodeError):
jpayne@7:     """ Base exception for all IDNA-encoding related problems """
jpayne@7:     pass
jpayne@7: 
jpayne@7: 
jpayne@7: class IDNABidiError(IDNAError):
jpayne@7:     """ Exception when bidirectional requirements are not satisfied """
jpayne@7:     pass
jpayne@7: 
jpayne@7: 
jpayne@7: class InvalidCodepoint(IDNAError):
jpayne@7:     """ Exception when a disallowed or unallocated codepoint is used """
jpayne@7:     pass
jpayne@7: 
jpayne@7: 
jpayne@7: class InvalidCodepointContext(IDNAError):
jpayne@7:     """ Exception when the codepoint is not valid in the context it is used """
jpayne@7:     pass
jpayne@7: 
jpayne@7: 
jpayne@7: def _combining_class(cp: int) -> int:
jpayne@7:     v = unicodedata.combining(chr(cp))
jpayne@7:     if v == 0:
jpayne@7:         if not unicodedata.name(chr(cp)):
jpayne@7:             raise ValueError('Unknown character in unicodedata')
jpayne@7:     return v
jpayne@7: 
jpayne@7: def _is_script(cp: str, script: str) -> bool:
jpayne@7:     return intranges_contain(ord(cp), idnadata.scripts[script])
jpayne@7: 
jpayne@7: def _punycode(s: str) -> bytes:
jpayne@7:     return s.encode('punycode')
jpayne@7: 
jpayne@7: def _unot(s: int) -> str:
jpayne@7:     return 'U+{:04X}'.format(s)
jpayne@7: 
jpayne@7: 
jpayne@7: def valid_label_length(label: Union[bytes, str]) -> bool:
jpayne@7:     if len(label) > 63:
jpayne@7:         return False
jpayne@7:     return True
jpayne@7: 
jpayne@7: 
jpayne@7: def valid_string_length(label: Union[bytes, str], trailing_dot: bool) -> bool:
jpayne@7:     if len(label) > (254 if trailing_dot else 253):
jpayne@7:         return False
jpayne@7:     return True
jpayne@7: 
jpayne@7: 
jpayne@7: def check_bidi(label: str, check_ltr: bool = False) -> bool:
jpayne@7:     # Bidi rules should only be applied if string contains RTL characters
jpayne@7:     bidi_label = False
jpayne@7:     for (idx, cp) in enumerate(label, 1):
jpayne@7:         direction = unicodedata.bidirectional(cp)
jpayne@7:         if direction == '':
jpayne@7:             # String likely comes from a newer version of Unicode
jpayne@7:             raise IDNABidiError('Unknown directionality in label {} at position {}'.format(repr(label), idx))
jpayne@7:         if direction in ['R', 'AL', 'AN']:
jpayne@7:             bidi_label = True
jpayne@7:     if not bidi_label and not check_ltr:
jpayne@7:         return True
jpayne@7: 
jpayne@7:     # Bidi rule 1
jpayne@7:     direction = unicodedata.bidirectional(label[0])
jpayne@7:     if direction in ['R', 'AL']:
jpayne@7:         rtl = True
jpayne@7:     elif direction == 'L':
jpayne@7:         rtl = False
jpayne@7:     else:
jpayne@7:         raise IDNABidiError('First codepoint in label {} must be directionality L, R or AL'.format(repr(label)))
jpayne@7: 
jpayne@7:     valid_ending = False
jpayne@7:     number_type = None  # type: Optional[str]
jpayne@7:     for (idx, cp) in enumerate(label, 1):
jpayne@7:         direction = unicodedata.bidirectional(cp)
jpayne@7: 
jpayne@7:         if rtl:
jpayne@7:             # Bidi rule 2
jpayne@7:             if not direction in ['R', 'AL', 'AN', 'EN', 'ES', 'CS', 'ET', 'ON', 'BN', 'NSM']:
jpayne@7:                 raise IDNABidiError('Invalid direction for codepoint at position {} in a right-to-left label'.format(idx))
jpayne@7:             # Bidi rule 3
jpayne@7:             if direction in ['R', 'AL', 'EN', 'AN']:
jpayne@7:                 valid_ending = True
jpayne@7:             elif direction != 'NSM':
jpayne@7:                 valid_ending = False
jpayne@7:             # Bidi rule 4
jpayne@7:             if direction in ['AN', 'EN']:
jpayne@7:                 if not number_type:
jpayne@7:                     number_type = direction
jpayne@7:                 else:
jpayne@7:                     if number_type != direction:
jpayne@7:                         raise IDNABidiError('Can not mix numeral types in a right-to-left label')
jpayne@7:         else:
jpayne@7:             # Bidi rule 5
jpayne@7:             if not direction in ['L', 'EN', 'ES', 'CS', 'ET', 'ON', 'BN', 'NSM']:
jpayne@7:                 raise IDNABidiError('Invalid direction for codepoint at position {} in a left-to-right label'.format(idx))
jpayne@7:             # Bidi rule 6
jpayne@7:             if direction in ['L', 'EN']:
jpayne@7:                 valid_ending = True
jpayne@7:             elif direction != 'NSM':
jpayne@7:                 valid_ending = False
jpayne@7: 
jpayne@7:     if not valid_ending:
jpayne@7:         raise IDNABidiError('Label ends with illegal codepoint directionality')
jpayne@7: 
jpayne@7:     return True
jpayne@7: 
jpayne@7: 
jpayne@7: def check_initial_combiner(label: str) -> bool:
jpayne@7:     if unicodedata.category(label[0])[0] == 'M':
jpayne@7:         raise IDNAError('Label begins with an illegal combining character')
jpayne@7:     return True
jpayne@7: 
jpayne@7: 
jpayne@7: def check_hyphen_ok(label: str) -> bool:
jpayne@7:     if label[2:4] == '--':
jpayne@7:         raise IDNAError('Label has disallowed hyphens in 3rd and 4th position')
jpayne@7:     if label[0] == '-' or label[-1] == '-':
jpayne@7:         raise IDNAError('Label must not start or end with a hyphen')
jpayne@7:     return True
jpayne@7: 
jpayne@7: 
jpayne@7: def check_nfc(label: str) -> None:
jpayne@7:     if unicodedata.normalize('NFC', label) != label:
jpayne@7:         raise IDNAError('Label must be in Normalization Form C')
jpayne@7: 
jpayne@7: 
jpayne@7: def valid_contextj(label: str, pos: int) -> bool:
jpayne@7:     cp_value = ord(label[pos])
jpayne@7: 
jpayne@7:     if cp_value == 0x200c:
jpayne@7: 
jpayne@7:         if pos > 0:
jpayne@7:             if _combining_class(ord(label[pos - 1])) == _virama_combining_class:
jpayne@7:                 return True
jpayne@7: 
jpayne@7:         ok = False
jpayne@7:         for i in range(pos-1, -1, -1):
jpayne@7:             joining_type = idnadata.joining_types.get(ord(label[i]))
jpayne@7:             if joining_type == ord('T'):
jpayne@7:                 continue
jpayne@7:             elif joining_type in [ord('L'), ord('D')]:
jpayne@7:                 ok = True
jpayne@7:                 break
jpayne@7:             else:
jpayne@7:                 break
jpayne@7: 
jpayne@7:         if not ok:
jpayne@7:             return False
jpayne@7: 
jpayne@7:         ok = False
jpayne@7:         for i in range(pos+1, len(label)):
jpayne@7:             joining_type = idnadata.joining_types.get(ord(label[i]))
jpayne@7:             if joining_type == ord('T'):
jpayne@7:                 continue
jpayne@7:             elif joining_type in [ord('R'), ord('D')]:
jpayne@7:                 ok = True
jpayne@7:                 break
jpayne@7:             else:
jpayne@7:                 break
jpayne@7:         return ok
jpayne@7: 
jpayne@7:     if cp_value == 0x200d:
jpayne@7: 
jpayne@7:         if pos > 0:
jpayne@7:             if _combining_class(ord(label[pos - 1])) == _virama_combining_class:
jpayne@7:                 return True
jpayne@7:         return False
jpayne@7: 
jpayne@7:     else:
jpayne@7: 
jpayne@7:         return False
jpayne@7: 
jpayne@7: 
jpayne@7: def valid_contexto(label: str, pos: int, exception: bool = False) -> bool:
jpayne@7:     cp_value = ord(label[pos])
jpayne@7: 
jpayne@7:     if cp_value == 0x00b7:
jpayne@7:         if 0 < pos < len(label)-1:
jpayne@7:             if ord(label[pos - 1]) == 0x006c and ord(label[pos + 1]) == 0x006c:
jpayne@7:                 return True
jpayne@7:         return False
jpayne@7: 
jpayne@7:     elif cp_value == 0x0375:
jpayne@7:         if pos < len(label)-1 and len(label) > 1:
jpayne@7:             return _is_script(label[pos + 1], 'Greek')
jpayne@7:         return False
jpayne@7: 
jpayne@7:     elif cp_value == 0x05f3 or cp_value == 0x05f4:
jpayne@7:         if pos > 0:
jpayne@7:             return _is_script(label[pos - 1], 'Hebrew')
jpayne@7:         return False
jpayne@7: 
jpayne@7:     elif cp_value == 0x30fb:
jpayne@7:         for cp in label:
jpayne@7:             if cp == '\u30fb':
jpayne@7:                 continue
jpayne@7:             if _is_script(cp, 'Hiragana') or _is_script(cp, 'Katakana') or _is_script(cp, 'Han'):
jpayne@7:                 return True
jpayne@7:         return False
jpayne@7: 
jpayne@7:     elif 0x660 <= cp_value <= 0x669:
jpayne@7:         for cp in label:
jpayne@7:             if 0x6f0 <= ord(cp) <= 0x06f9:
jpayne@7:                 return False
jpayne@7:         return True
jpayne@7: 
jpayne@7:     elif 0x6f0 <= cp_value <= 0x6f9:
jpayne@7:         for cp in label:
jpayne@7:             if 0x660 <= ord(cp) <= 0x0669:
jpayne@7:                 return False
jpayne@7:         return True
jpayne@7: 
jpayne@7:     return False
jpayne@7: 
jpayne@7: 
jpayne@7: def check_label(label: Union[str, bytes, bytearray]) -> None:
jpayne@7:     if isinstance(label, (bytes, bytearray)):
jpayne@7:         label = label.decode('utf-8')
jpayne@7:     if len(label) == 0:
jpayne@7:         raise IDNAError('Empty Label')
jpayne@7: 
jpayne@7:     check_nfc(label)
jpayne@7:     check_hyphen_ok(label)
jpayne@7:     check_initial_combiner(label)
jpayne@7: 
jpayne@7:     for (pos, cp) in enumerate(label):
jpayne@7:         cp_value = ord(cp)
jpayne@7:         if intranges_contain(cp_value, idnadata.codepoint_classes['PVALID']):
jpayne@7:             continue
jpayne@7:         elif intranges_contain(cp_value, idnadata.codepoint_classes['CONTEXTJ']):
jpayne@7:             if not valid_contextj(label, pos):
jpayne@7:                 raise InvalidCodepointContext('Joiner {} not allowed at position {} in {}'.format(
jpayne@7:                     _unot(cp_value), pos+1, repr(label)))
jpayne@7:         elif intranges_contain(cp_value, idnadata.codepoint_classes['CONTEXTO']):
jpayne@7:             if not valid_contexto(label, pos):
jpayne@7:                 raise InvalidCodepointContext('Codepoint {} not allowed at position {} in {}'.format(_unot(cp_value), pos+1, repr(label)))
jpayne@7:         else:
jpayne@7:             raise InvalidCodepoint('Codepoint {} at position {} of {} not allowed'.format(_unot(cp_value), pos+1, repr(label)))
jpayne@7: 
jpayne@7:     check_bidi(label)
jpayne@7: 
jpayne@7: 
jpayne@7: def alabel(label: str) -> bytes:
jpayne@7:     try:
jpayne@7:         label_bytes = label.encode('ascii')
jpayne@7:         ulabel(label_bytes)
jpayne@7:         if not valid_label_length(label_bytes):
jpayne@7:             raise IDNAError('Label too long')
jpayne@7:         return label_bytes
jpayne@7:     except UnicodeEncodeError:
jpayne@7:         pass
jpayne@7: 
jpayne@7:     check_label(label)
jpayne@7:     label_bytes = _alabel_prefix + _punycode(label)
jpayne@7: 
jpayne@7:     if not valid_label_length(label_bytes):
jpayne@7:         raise IDNAError('Label too long')
jpayne@7: 
jpayne@7:     return label_bytes
jpayne@7: 
jpayne@7: 
jpayne@7: def ulabel(label: Union[str, bytes, bytearray]) -> str:
jpayne@7:     if not isinstance(label, (bytes, bytearray)):
jpayne@7:         try:
jpayne@7:             label_bytes = label.encode('ascii')
jpayne@7:         except UnicodeEncodeError:
jpayne@7:             check_label(label)
jpayne@7:             return label
jpayne@7:     else:
jpayne@7:         label_bytes = label
jpayne@7: 
jpayne@7:     label_bytes = label_bytes.lower()
jpayne@7:     if label_bytes.startswith(_alabel_prefix):
jpayne@7:         label_bytes = label_bytes[len(_alabel_prefix):]
jpayne@7:         if not label_bytes:
jpayne@7:             raise IDNAError('Malformed A-label, no Punycode eligible content found')
jpayne@7:         if label_bytes.decode('ascii')[-1] == '-':
jpayne@7:             raise IDNAError('A-label must not end with a hyphen')
jpayne@7:     else:
jpayne@7:         check_label(label_bytes)
jpayne@7:         return label_bytes.decode('ascii')
jpayne@7: 
jpayne@7:     try:
jpayne@7:         label = label_bytes.decode('punycode')
jpayne@7:     except UnicodeError:
jpayne@7:         raise IDNAError('Invalid A-label')
jpayne@7:     check_label(label)
jpayne@7:     return label
jpayne@7: 
jpayne@7: 
jpayne@7: def uts46_remap(domain: str, std3_rules: bool = True, transitional: bool = False) -> str:
jpayne@7:     """Re-map the characters in the string according to UTS46 processing."""
jpayne@7:     from .uts46data import uts46data
jpayne@7:     output = ''
jpayne@7: 
jpayne@7:     for pos, char in enumerate(domain):
jpayne@7:         code_point = ord(char)
jpayne@7:         try:
jpayne@7:             uts46row = uts46data[code_point if code_point < 256 else
jpayne@7:                 bisect.bisect_left(uts46data, (code_point, 'Z')) - 1]
jpayne@7:             status = uts46row[1]
jpayne@7:             replacement = None  # type: Optional[str]
jpayne@7:             if len(uts46row) == 3:
jpayne@7:                 replacement = uts46row[2]
jpayne@7:             if (status == 'V' or
jpayne@7:                     (status == 'D' and not transitional) or
jpayne@7:                     (status == '3' and not std3_rules and replacement is None)):
jpayne@7:                 output += char
jpayne@7:             elif replacement is not None and (status == 'M' or
jpayne@7:                     (status == '3' and not std3_rules) or
jpayne@7:                     (status == 'D' and transitional)):
jpayne@7:                 output += replacement
jpayne@7:             elif status != 'I':
jpayne@7:                 raise IndexError()
jpayne@7:         except IndexError:
jpayne@7:             raise InvalidCodepoint(
jpayne@7:                 'Codepoint {} not allowed at position {} in {}'.format(
jpayne@7:                 _unot(code_point), pos + 1, repr(domain)))
jpayne@7: 
jpayne@7:     return unicodedata.normalize('NFC', output)
jpayne@7: 
jpayne@7: 
jpayne@7: def encode(s: Union[str, bytes, bytearray], strict: bool = False, uts46: bool = False, std3_rules: bool = False, transitional: bool = False) -> bytes:
jpayne@7:     if not isinstance(s, str):
jpayne@7:         try:
jpayne@7:             s = str(s, 'ascii')
jpayne@7:         except UnicodeDecodeError:
jpayne@7:             raise IDNAError('should pass a unicode string to the function rather than a byte string.')
jpayne@7:     if uts46:
jpayne@7:         s = uts46_remap(s, std3_rules, transitional)
jpayne@7:     trailing_dot = False
jpayne@7:     result = []
jpayne@7:     if strict:
jpayne@7:         labels = s.split('.')
jpayne@7:     else:
jpayne@7:         labels = _unicode_dots_re.split(s)
jpayne@7:     if not labels or labels == ['']:
jpayne@7:         raise IDNAError('Empty domain')
jpayne@7:     if labels[-1] == '':
jpayne@7:         del labels[-1]
jpayne@7:         trailing_dot = True
jpayne@7:     for label in labels:
jpayne@7:         s = alabel(label)
jpayne@7:         if s:
jpayne@7:             result.append(s)
jpayne@7:         else:
jpayne@7:             raise IDNAError('Empty label')
jpayne@7:     if trailing_dot:
jpayne@7:         result.append(b'')
jpayne@7:     s = b'.'.join(result)
jpayne@7:     if not valid_string_length(s, trailing_dot):
jpayne@7:         raise IDNAError('Domain too long')
jpayne@7:     return s
jpayne@7: 
jpayne@7: 
jpayne@7: def decode(s: Union[str, bytes, bytearray], strict: bool = False, uts46: bool = False, std3_rules: bool = False) -> str:
jpayne@7:     try:
jpayne@7:         if not isinstance(s, str):
jpayne@7:             s = str(s, 'ascii')
jpayne@7:     except UnicodeDecodeError:
jpayne@7:         raise IDNAError('Invalid ASCII in A-label')
jpayne@7:     if uts46:
jpayne@7:         s = uts46_remap(s, std3_rules, False)
jpayne@7:     trailing_dot = False
jpayne@7:     result = []
jpayne@7:     if not strict:
jpayne@7:         labels = _unicode_dots_re.split(s)
jpayne@7:     else:
jpayne@7:         labels = s.split('.')
jpayne@7:     if not labels or labels == ['']:
jpayne@7:         raise IDNAError('Empty domain')
jpayne@7:     if not labels[-1]:
jpayne@7:         del labels[-1]
jpayne@7:         trailing_dot = True
jpayne@7:     for label in labels:
jpayne@7:         s = ulabel(label)
jpayne@7:         if s:
jpayne@7:             result.append(s)
jpayne@7:         else:
jpayne@7:             raise IDNAError('Empty label')
jpayne@7:     if trailing_dot:
jpayne@7:         result.append('')
jpayne@7:     return '.'.join(result)