X-Git-Url: http://www.git.cypherpunks.ru/?a=blobdiff_plain;f=pyderasn.py;h=50faa0be870ced74ab9d33afb4ac0f4008846421;hb=97fe8b8eae995499566e3ae26e2e941e633997bc;hp=3dd3ea9c151a3f1ca6f571a7765614e1c5bf5b91;hpb=f928d1e65eeff56c2b5d511567e6b52417195f3b;p=pyderasn.git diff --git a/pyderasn.py b/pyderasn.py index 3dd3ea9..50faa0b 100755 --- a/pyderasn.py +++ b/pyderasn.py @@ -1,7 +1,7 @@ #!/usr/bin/env python # coding: utf-8 # PyDERASN -- Python ASN.1 DER/BER codec with abstract structures -# Copyright (C) 2017-2018 Sergey Matveev +# Copyright (C) 2017-2019 Sergey Matveev # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Lesser General Public License as @@ -382,15 +382,22 @@ DER. But you can optionally enable BER decoding with setting ``bered`` :ref:`context ` argument to True. Indefinite lengths and constructed primitive types should be parsed successfully. -* If object is encoded in BER form (not the DER one), then ``bered`` +* If object is encoded in BER form (not the DER one), then ``ber_encoded`` attribute is set to True. Only ``BOOLEAN``, ``BIT STRING``, ``OCTET - STRING``, ``SEQUENCE``, ``SET`` can contain it. + STRING``, ``OBJECT IDENTIFIER``, ``SEQUENCE``, ``SET``, ``SET OF`` + can contain it. * If object has an indefinite length encoding, then its ``lenindef`` attribute is set to True. Only ``BIT STRING``, ``OCTET STRING``, ``SEQUENCE``, ``SET``, ``SEQUENCE OF``, ``SET OF``, ``ANY`` can contain it. * If object has an indefinite length encoded explicit tag, then ``expl_lenindef`` is set to True. +* If object has either any of BER-related encoding (explicit tag + indefinite length, object's indefinite length, BER-encoding) or any + underlying component has that kind of encoding, then ``bered`` + attribute is set to True. For example SignedData CMS can have + ``ContentInfo:content:signerInfos:*`` ``bered`` value set to True, but + ``ContentInfo:content:signerInfos:*:signedAttrs`` won't. EOC (end-of-contents) token's length is taken in advance in object's value length. @@ -509,6 +516,7 @@ Various ------- .. autofunction:: pyderasn.abs_decode_path +.. autofunction:: pyderasn.colonize_hex .. autofunction:: pyderasn.hexenc .. autofunction:: pyderasn.hexdec .. autofunction:: pyderasn.tag_encode @@ -533,9 +541,11 @@ from codecs import getdecoder from codecs import getencoder from collections import namedtuple from collections import OrderedDict +from copy import copy from datetime import datetime from math import ceil from os import environ +from string import ascii_letters from string import digits from six import add_metaclass @@ -545,15 +555,18 @@ from six import indexbytes from six import int2byte from six import integer_types from six import iterbytes +from six import iteritems +from six import itervalues from six import PY2 from six import string_types from six import text_type +from six import unichr as six_unichr from six.moves import xrange as six_xrange try: from termcolor import colored -except ImportError: +except ImportError: # pragma: no cover def colored(what, *args): return what @@ -635,7 +648,11 @@ LENINDEF_PP_CHAR = "I" if PY2 else "∞" # Errors ######################################################################## -class DecodeError(Exception): +class ASN1Error(ValueError): + pass + + +class DecodeError(ASN1Error): def __init__(self, msg="", klass=None, decode_path=(), offset=0): """ :param str msg: reason of decode failing @@ -690,7 +707,7 @@ class InvalidOID(DecodeError): pass -class ObjUnknown(ValueError): +class ObjUnknown(ASN1Error): def __init__(self, name): super(ObjUnknown, self).__init__() self.name = name @@ -702,7 +719,7 @@ class ObjUnknown(ValueError): return "%s(%s)" % (self.__class__.__name__, self) -class ObjNotReady(ValueError): +class ObjNotReady(ASN1Error): def __init__(self, name): super(ObjNotReady, self).__init__() self.name = name @@ -714,7 +731,7 @@ class ObjNotReady(ValueError): return "%s(%s)" % (self.__class__.__name__, self) -class InvalidValueType(ValueError): +class InvalidValueType(ASN1Error): def __init__(self, expected_types): super(InvalidValueType, self).__init__() self.expected_types = expected_types @@ -728,7 +745,7 @@ class InvalidValueType(ValueError): return "%s(%s)" % (self.__class__.__name__, self) -class BoundsError(ValueError): +class BoundsError(ASN1Error): def __init__(self, bound_min, value, bound_max): super(BoundsError, self).__init__() self.bound_min = bound_min @@ -922,7 +939,7 @@ class Obj(object): "vlen", "expl_lenindef", "lenindef", - "bered", + "ber_encoded", ) def __init__( @@ -944,7 +961,7 @@ class Obj(object): self.default = None self.expl_lenindef = False self.lenindef = False - self.bered = False + self.ber_encoded = False @property def ready(self): # pragma: no cover @@ -956,6 +973,12 @@ class Obj(object): if not self.ready: raise ObjNotReady(self.__class__.__name__) + @property + def bered(self): + """Is either object or any elements inside is BER encoded? + """ + return self.expl_lenindef or self.lenindef or self.ber_encoded + @property def decoded(self): """Is object decoded? @@ -1010,6 +1033,7 @@ class Obj(object): decode_path=(), ctx=None, tag_only=False, + _ctx_immutable=True, ): """Decode the data @@ -1017,14 +1041,17 @@ class Obj(object): :param int offset: initial data's offset :param bool leavemm: do we need to leave memoryview of remaining data as is, or convert it to bytes otherwise - :param ctx: optional :ref:`context ` governing decoding process. + :param ctx: optional :ref:`context ` governing decoding process :param tag_only: decode only the tag, without length and contents (used only in Choice and Set structures, trying to determine if tag satisfies the scheme) + :param _ctx_immutable: do we need to copy ``ctx`` before using it :returns: (Obj, remaining data) """ if ctx is None: ctx = {} + elif _ctx_immutable: + ctx = copy(ctx) tlv = memoryview(data) if self._expl is None: result = self._decode( @@ -1072,7 +1099,7 @@ class Obj(object): ctx=ctx, tag_only=tag_only, ) - if tag_only: + if tag_only: # pragma: no cover return obj, tail = result eoc_expected, tail = tail[:EOC_LEN], tail[EOC_LEN:] @@ -1107,7 +1134,7 @@ class Obj(object): ctx=ctx, tag_only=tag_only, ) - if tag_only: + if tag_only: # pragma: no cover return obj, tail = result if obj.tlvlen < l and not ctx.get("allow_expl_oob", False): @@ -1158,7 +1185,10 @@ class Obj(object): return self.expl_tlvlen if self.expled else self.tlvlen def pps_lenindef(self, decode_path): - if self.lenindef: + if self.lenindef and not ( + getattr(self, "defined", None) is not None and + self.defined[1].lenindef + ): yield _pp( asn1_type_name="EOC", obj_name="", @@ -1170,6 +1200,7 @@ class Obj(object): tlen=1, llen=1, vlen=0, + ber_encoded=True, bered=True, ) if self.expl_lenindef: @@ -1181,6 +1212,7 @@ class Obj(object): tlen=1, llen=1, vlen=0, + ber_encoded=True, bered=True, ) @@ -1213,6 +1245,7 @@ class DecodePathDefBy(object): ######################################################################## PP = namedtuple("PP", ( + "obj", "asn1_type_name", "obj_name", "decode_path", @@ -1232,11 +1265,13 @@ PP = namedtuple("PP", ( "expl_vlen", "expl_lenindef", "lenindef", + "ber_encoded", "bered", )) def _pp( + obj=None, asn1_type_name="unknown", obj_name="unknown", decode_path=(), @@ -1256,9 +1291,11 @@ def _pp( expl_vlen=None, expl_lenindef=False, lenindef=False, + ber_encoded=False, bered=False, ): return PP( + obj, asn1_type_name, obj_name, decode_path, @@ -1278,6 +1315,7 @@ def _pp( expl_vlen, expl_lenindef, lenindef, + ber_encoded, bered, ) @@ -1286,6 +1324,12 @@ def _colourize(what, colour, with_colours, attrs=("bold",)): return colored(what, colour, attrs=attrs) if with_colours else what +def colonize_hex(hexed): + """Separate hexadecimal string with colons + """ + return ":".join(hexed[i:i + 2] for i in six_xrange(0, len(hexed), 2)) + + def pp_console_row( pp, oids=None, @@ -1305,7 +1349,9 @@ def pp_console_row( ), LENINDEF_PP_CHAR if pp.expl_lenindef else " ", ) - cols.append(_colourize(col, "red", with_colours, ())) + col = _colourize(col, "red", with_colours, ()) + col += _colourize("B", "red", with_colours) if pp.bered else " " + cols.append(col) col = "[%d,%d,%4d]%s" % ( pp.tlen, pp.llen, @@ -1342,7 +1388,7 @@ def pp_console_row( cols.append(_colourize(col, "blue", with_colours)) if pp.asn1_type_name.replace(" ", "") != pp.obj_name.upper(): cols.append(_colourize(pp.obj_name, "magenta", with_colours)) - if pp.bered: + if pp.ber_encoded: cols.append(_colourize("BER", "red", with_colours)) cols.append(_colourize(pp.asn1_type_name, "cyan", with_colours)) if pp.value is not None: @@ -1354,6 +1400,15 @@ def pp_console_row( value in oids ): cols.append(_colourize("(%s)" % oids[value], "green", with_colours)) + if pp.asn1_type_name == Integer.asn1_type_name: + hex_repr = hex(int(pp.obj._value))[2:].upper() + if len(hex_repr) % 2 != 0: + hex_repr = "0" + hex_repr + cols.append(_colourize( + "(%s)" % colonize_hex(hex_repr), + "green", + with_colours, + )) if with_blob: if isinstance(pp.blob, binary_type): cols.append(hexenc(pp.blob)) @@ -1373,17 +1428,15 @@ def pp_console_row( def pp_console_blob(pp, decode_path_len_decrease=0): - cols = [" " * len("XXXXXYYZ [X,X,XXXX]Z")] + cols = [" " * len("XXXXXYYZZ [X,X,XXXX]Z")] decode_path_len = len(pp.decode_path) - decode_path_len_decrease if decode_path_len > 0: cols.append(" ." * (decode_path_len + 1)) if isinstance(pp.blob, binary_type): blob = hexenc(pp.blob).upper() - for i in range(0, len(blob), 32): + for i in six_xrange(0, len(blob), 32): chunk = blob[i:i + 32] - yield " ".join(cols + [":".join( - chunk[j:j + 2] for j in range(0, len(chunk), 2) - )]) + yield " ".join(cols + [colonize_hex(chunk)]) elif isinstance(pp.blob, tuple): yield " ".join(cols + [", ".join(pp.blob)]) @@ -1498,10 +1551,10 @@ class Boolean(Obj): self._value = default def _value_sanitize(self, value): - if issubclass(value.__class__, Boolean): - return value._value if isinstance(value, bool): return value + if issubclass(value.__class__, Boolean): + return value._value raise InvalidValueType((self.__class__, bool)) @property @@ -1518,6 +1571,9 @@ class Boolean(Obj): obj.offset = self.offset obj.llen = self.llen obj.vlen = self.vlen + obj.expl_lenindef = self.expl_lenindef + obj.lenindef = self.lenindef + obj.ber_encoded = self.ber_encoded return obj def __nonzero__(self): @@ -1605,14 +1661,14 @@ class Boolean(Obj): offset=offset, ) first_octet = byte2int(v) - bered = False + ber_encoded = False if first_octet == 0: value = False elif first_octet == 0xFF: value = True elif ctx.get("bered", False): value = True - bered = True + ber_encoded = True else: raise DecodeError( "unacceptable Boolean value", @@ -1628,7 +1684,7 @@ class Boolean(Obj): optional=self.optional, _decoded=(offset, 1, 1), ) - obj.bered = bered + obj.ber_encoded = ber_encoded return obj, v[1:] def __repr__(self): @@ -1636,6 +1692,7 @@ class Boolean(Obj): def pps(self, decode_path=()): yield _pp( + obj=self, asn1_type_name=self.asn1_type_name, obj_name=self.__class__.__name__, decode_path=decode_path, @@ -1653,6 +1710,7 @@ class Boolean(Obj): expl_llen=self.expl_llen if self.expled else None, expl_vlen=self.expl_vlen if self.expled else None, expl_lenindef=self.expl_lenindef, + ber_encoded=self.ber_encoded, bered=self.bered, ) for pp in self.pps_lenindef(decode_path): @@ -1742,10 +1800,10 @@ class Integer(Obj): self._value = default def _value_sanitize(self, value): - if issubclass(value.__class__, Integer): - value = value._value - elif isinstance(value, integer_types): + if isinstance(value, integer_types): pass + elif issubclass(value.__class__, Integer): + value = value._value elif isinstance(value, str): value = self.specs.get(value) if value is None: @@ -1772,6 +1830,9 @@ class Integer(Obj): obj.offset = self.offset obj.llen = self.llen obj.vlen = self.vlen + obj.expl_lenindef = self.expl_lenindef + obj.lenindef = self.lenindef + obj.ber_encoded = self.ber_encoded return obj def __int__(self): @@ -1802,7 +1863,7 @@ class Integer(Obj): @property def named(self): - for name, value in self.specs.items(): + for name, value in iteritems(self.specs): if value == self._value: return name @@ -1962,6 +2023,7 @@ class Integer(Obj): def pps(self, decode_path=()): yield _pp( + obj=self, asn1_type_name=self.asn1_type_name, obj_name=self.__class__.__name__, decode_path=decode_path, @@ -1979,11 +2041,15 @@ class Integer(Obj): expl_llen=self.expl_llen if self.expled else None, expl_vlen=self.expl_vlen if self.expled else None, expl_lenindef=self.expl_lenindef, + bered=self.bered, ) for pp in self.pps_lenindef(decode_path): yield pp +SET01 = frozenset(("0", "1")) + + class BitString(Obj): """``BIT STRING`` bit string type @@ -2089,8 +2155,6 @@ class BitString(Obj): return bit_len, bytes(octets) def _value_sanitize(self, value): - if issubclass(value.__class__, BitString): - return value._value if isinstance(value, (string_types, binary_type)): if ( isinstance(value, string_types) and @@ -2098,7 +2162,7 @@ class BitString(Obj): ): if value.endswith("'B"): value = value[1:-2] - if not set(value) <= set(("0", "1")): + if not frozenset(value) <= SET01: raise ValueError("B's coding contains unacceptable chars") return self._bits2octets(value) elif value.endswith("'H"): @@ -2126,11 +2190,13 @@ class BitString(Obj): bits.append(bit) if len(bits) == 0: return self._bits2octets("") - bits = set(bits) + bits = frozenset(bits) return self._bits2octets("".join( ("1" if bit in bits else "0") for bit in six_xrange(max(bits) + 1) )) + if issubclass(value.__class__, BitString): + return value._value raise InvalidValueType((self.__class__, binary_type, string_types)) @property @@ -2150,6 +2216,9 @@ class BitString(Obj): obj.offset = self.offset obj.llen = self.llen obj.vlen = self.vlen + obj.expl_lenindef = self.expl_lenindef + obj.lenindef = self.lenindef + obj.ber_encoded = self.ber_encoded return obj def __iter__(self): @@ -2179,7 +2248,7 @@ class BitString(Obj): @property def named(self): - return [name for name, bit in self.specs.items() if self[bit]] + return [name for name, bit in iteritems(self.specs) if self[bit]] def __call__( self, @@ -2293,7 +2362,7 @@ class BitString(Obj): offset=offset, ) if t == self.tag: - if tag_only: + if tag_only: # pragma: no cover return return self._decode_chunk(lv, offset, decode_path, ctx) if t == self.tag_constructed: @@ -2304,7 +2373,7 @@ class BitString(Obj): decode_path=decode_path, offset=offset, ) - if tag_only: + if tag_only: # pragma: no cover return lenindef = False try: @@ -2358,6 +2427,7 @@ class BitString(Obj): decode_path=sub_decode_path, leavemm=True, ctx=ctx, + _ctx_immutable=False, ) except TagMismatch: raise DecodeError( @@ -2402,7 +2472,7 @@ class BitString(Obj): _decoded=(offset, llen, vlen + (EOC_LEN if lenindef else 0)), ) obj.lenindef = lenindef - obj.bered = True + obj.ber_encoded = True return obj, (v[EOC_LEN:] if lenindef else v) raise TagMismatch( klass=self.__class__, @@ -2422,6 +2492,7 @@ class BitString(Obj): if len(self.specs) > 0: blob = tuple(self.named) yield _pp( + obj=self, asn1_type_name=self.asn1_type_name, obj_name=self.__class__.__name__, decode_path=decode_path, @@ -2441,6 +2512,7 @@ class BitString(Obj): expl_vlen=self.expl_vlen if self.expled else None, expl_lenindef=self.expl_lenindef, lenindef=self.lenindef, + ber_encoded=self.ber_encoded, bered=self.bered, ) defined_by, defined = self.defined or (None, None) @@ -2533,10 +2605,10 @@ class OctetString(Obj): ) def _value_sanitize(self, value): - if issubclass(value.__class__, OctetString): - value = value._value - elif isinstance(value, binary_type): + if isinstance(value, binary_type): pass + elif issubclass(value.__class__, OctetString): + value = value._value else: raise InvalidValueType((self.__class__, bytes)) if not self._bound_min <= len(value) <= self._bound_max: @@ -2559,6 +2631,9 @@ class OctetString(Obj): obj.offset = self.offset obj.llen = self.llen obj.vlen = self.vlen + obj.expl_lenindef = self.expl_lenindef + obj.lenindef = self.lenindef + obj.ber_encoded = self.ber_encoded return obj def __bytes__(self): @@ -2721,6 +2796,7 @@ class OctetString(Obj): decode_path=sub_decode_path, leavemm=True, ctx=ctx, + _ctx_immutable=False, ) except TagMismatch: raise DecodeError( @@ -2758,7 +2834,7 @@ class OctetString(Obj): offset=offset, ) obj.lenindef = lenindef - obj.bered = True + obj.ber_encoded = True return obj, (v[EOC_LEN:] if lenindef else v) raise TagMismatch( klass=self.__class__, @@ -2771,6 +2847,7 @@ class OctetString(Obj): def pps(self, decode_path=()): yield _pp( + obj=self, asn1_type_name=self.asn1_type_name, obj_name=self.__class__.__name__, decode_path=decode_path, @@ -2790,6 +2867,7 @@ class OctetString(Obj): expl_vlen=self.expl_vlen if self.expled else None, expl_lenindef=self.expl_lenindef, lenindef=self.lenindef, + ber_encoded=self.ber_encoded, bered=self.bered, ) defined_by, defined = self.defined or (None, None) @@ -2842,6 +2920,9 @@ class Null(Obj): obj.offset = self.offset obj.llen = self.llen obj.vlen = self.vlen + obj.expl_lenindef = self.expl_lenindef + obj.lenindef = self.lenindef + obj.ber_encoded = self.ber_encoded return obj def __eq__(self, their): @@ -2884,7 +2965,7 @@ class Null(Obj): decode_path=decode_path, offset=offset, ) - if tag_only: + if tag_only: # pragma: no cover return try: l, _, v = len_decode(lv) @@ -2915,6 +2996,7 @@ class Null(Obj): def pps(self, decode_path=()): yield _pp( + obj=self, asn1_type_name=self.asn1_type_name, obj_name=self.__class__.__name__, decode_path=decode_path, @@ -2930,6 +3012,7 @@ class Null(Obj): expl_llen=self.expl_llen if self.expled else None, expl_vlen=self.expl_vlen if self.expled else None, expl_lenindef=self.expl_lenindef, + bered=self.bered, ) for pp in self.pps_lenindef(decode_path): yield pp @@ -3050,6 +3133,9 @@ class ObjectIdentifier(Obj): obj.offset = self.offset obj.llen = self.llen obj.vlen = self.vlen + obj.expl_lenindef = self.expl_lenindef + obj.lenindef = self.lenindef + obj.ber_encoded = self.ber_encoded return obj def __iter__(self): @@ -3134,7 +3220,7 @@ class ObjectIdentifier(Obj): decode_path=decode_path, offset=offset, ) - if tag_only: + if tag_only: # pragma: no cover return try: l, llen, v = len_decode(lv) @@ -3161,11 +3247,17 @@ class ObjectIdentifier(Obj): ) v, tail = v[:l], v[l:] arcs = [] + ber_encoded = False while len(v) > 0: i = 0 arc = 0 while True: octet = indexbytes(v, i) + if i == 0 and octet == 0x80: + if ctx.get("bered", False): + ber_encoded = True + else: + raise DecodeError("non normalized arc encoding") arc = (arc << 7) | (octet & 0x7F) if octet & 0x80 == 0: arcs.append(arc) @@ -3197,6 +3289,8 @@ class ObjectIdentifier(Obj): optional=self.optional, _decoded=(offset, llen, l), ) + if ber_encoded: + obj.ber_encoded = True return obj, tail def __repr__(self): @@ -3204,6 +3298,7 @@ class ObjectIdentifier(Obj): def pps(self, decode_path=()): yield _pp( + obj=self, asn1_type_name=self.asn1_type_name, obj_name=self.__class__.__name__, decode_path=decode_path, @@ -3221,6 +3316,8 @@ class ObjectIdentifier(Obj): expl_llen=self.expl_llen if self.expled else None, expl_vlen=self.expl_vlen if self.expled else None, expl_lenindef=self.expl_lenindef, + ber_encoded=self.ber_encoded, + bered=self.bered, ) for pp in self.pps_lenindef(decode_path): yield pp @@ -3263,7 +3360,10 @@ class Enumerated(Integer): if isinstance(value, self.__class__): value = value._value elif isinstance(value, integer_types): - if value not in list(self.specs.values()): + for _value in itervalues(self.specs): + if _value == value: + break + else: raise DecodeError( "unknown integer value: %s" % value, klass=self.__class__, @@ -3288,6 +3388,9 @@ class Enumerated(Integer): obj.offset = self.offset obj.llen = self.llen obj.vlen = self.vlen + obj.expl_lenindef = self.expl_lenindef + obj.lenindef = self.lenindef + obj.ber_encoded = self.ber_encoded return obj def __call__( @@ -3429,6 +3532,7 @@ class CommonString(OctetString): if self.ready: value = hexenc(bytes(self)) if no_unicode else self.__unicode__() yield _pp( + obj=self, asn1_type_name=self.asn1_type_name, obj_name=self.__class__.__name__, decode_path=decode_path, @@ -3446,6 +3550,8 @@ class CommonString(OctetString): expl_llen=self.expl_llen if self.expled else None, expl_vlen=self.expl_vlen if self.expled else None, expl_lenindef=self.expl_lenindef, + ber_encoded=self.ber_encoded, + bered=self.bered, ) for pp in self.pps_lenindef(decode_path): yield pp @@ -3458,29 +3564,57 @@ class UTF8String(CommonString): asn1_type_name = "UTF8String" -class NumericString(CommonString): +class AllowableCharsMixin(object): + @property + def allowable_chars(self): + if PY2: + return self._allowable_chars + return frozenset(six_unichr(c) for c in self._allowable_chars) + + +class NumericString(AllowableCharsMixin, CommonString): """Numeric string - Its value is properly sanitized: only ASCII digits can be stored. + Its value is properly sanitized: only ASCII digits with spaces can + be stored. + + >>> NumericString().allowable_chars + set(['3', '4', '7', '5', '1', '0', '8', '9', ' ', '6', '2']) """ __slots__ = () tag_default = tag_encode(18) encoding = "ascii" asn1_type_name = "NumericString" - allowable_chars = set(digits.encode("ascii")) + _allowable_chars = frozenset(digits.encode("ascii") + b" ") def _value_sanitize(self, value): value = super(NumericString, self)._value_sanitize(value) - if not set(value) <= self.allowable_chars: + if not frozenset(value) <= self._allowable_chars: raise DecodeError("non-numeric value") return value -class PrintableString(CommonString): +class PrintableString(AllowableCharsMixin, CommonString): + """Printable string + + Its value is properly sanitized: see X.680 41.4 table 10. + + >>> PrintableString().allowable_chars + >>> set([' ', "'", ..., 'z']) + """ __slots__ = () tag_default = tag_encode(19) encoding = "ascii" asn1_type_name = "PrintableString" + _allowable_chars = frozenset( + (ascii_letters + digits + " '()+,-./:=?").encode("ascii") + ) + + def _value_sanitize(self, value): + value = super(PrintableString, self)._value_sanitize(value) + if not frozenset(value) <= self._allowable_chars: + raise DecodeError("non-printable value") + return value class TeletexString(CommonString): @@ -3574,10 +3708,6 @@ class UTCTime(CommonString): self._value = default def _value_sanitize(self, value): - if isinstance(value, self.__class__): - return value._value - if isinstance(value, datetime): - return value.strftime(self.fmt).encode("ascii") if isinstance(value, binary_type): try: value_decoded = value.decode("ascii") @@ -3591,6 +3721,10 @@ class UTCTime(CommonString): return value else: raise DecodeError("invalid UTCTime length") + if isinstance(value, self.__class__): + return value._value + if isinstance(value, datetime): + return value.strftime(self.fmt).encode("ascii") raise InvalidValueType((self.__class__, datetime)) def __eq__(self, their): @@ -3631,6 +3765,7 @@ class UTCTime(CommonString): def pps(self, decode_path=()): yield _pp( + obj=self, asn1_type_name=self.asn1_type_name, obj_name=self.__class__.__name__, decode_path=decode_path, @@ -3648,6 +3783,8 @@ class UTCTime(CommonString): expl_llen=self.expl_llen if self.expled else None, expl_vlen=self.expl_vlen if self.expled else None, expl_lenindef=self.expl_lenindef, + ber_encoded=self.ber_encoded, + bered=self.bered, ) for pp in self.pps_lenindef(decode_path): yield pp @@ -3673,12 +3810,6 @@ class GeneralizedTime(UTCTime): fmt_ms = "%Y%m%d%H%M%S.%fZ" def _value_sanitize(self, value): - if isinstance(value, self.__class__): - return value._value - if isinstance(value, datetime): - return value.strftime( - self.fmt_ms if value.microsecond > 0 else self.fmt - ).encode("ascii") if isinstance(value, binary_type): try: value_decoded = value.decode("ascii") @@ -3705,6 +3836,12 @@ class GeneralizedTime(UTCTime): "invalid GeneralizedTime length", klass=self.__class__, ) + if isinstance(value, self.__class__): + return value._value + if isinstance(value, datetime): + return value.strftime( + self.fmt_ms if value.microsecond > 0 else self.fmt + ).encode("ascii") raise InvalidValueType((self.__class__, datetime)) def todatetime(self): @@ -3830,8 +3967,6 @@ class Choice(Obj): self._value = default_obj.copy()._value def _value_sanitize(self, value): - if isinstance(value, self.__class__): - return value._value if isinstance(value, tuple) and len(value) == 2: choice, obj = value spec = self.specs.get(choice) @@ -3840,12 +3975,21 @@ class Choice(Obj): if not isinstance(obj, spec.__class__): raise InvalidValueType((spec,)) return (choice, spec(obj)) + if isinstance(value, self.__class__): + return value._value raise InvalidValueType((self.__class__, tuple)) @property def ready(self): return self._value is not None and self._value[1].ready + @property + def bered(self): + return self.expl_lenindef or ( + (self._value is not None) and + self._value[1].bered + ) + def copy(self): obj = self.__class__(schema=self.specs) obj._expl = self._expl @@ -3854,6 +3998,9 @@ class Choice(Obj): obj.offset = self.offset obj.llen = self.llen obj.vlen = self.vlen + obj.expl_lenindef = self.expl_lenindef + obj.lenindef = self.lenindef + obj.ber_encoded = self.ber_encoded value = self._value if value is not None: obj._value = (value[0], value[1].copy()) @@ -3925,7 +4072,7 @@ class Choice(Obj): return self._value[1].encode() def _decode(self, tlv, offset, decode_path, ctx, tag_only): - for choice, spec in self.specs.items(): + for choice, spec in iteritems(self.specs): sub_decode_path = decode_path + (choice,) try: spec.decode( @@ -3935,6 +4082,7 @@ class Choice(Obj): decode_path=sub_decode_path, ctx=ctx, tag_only=True, + _ctx_immutable=False, ) except TagMismatch: continue @@ -3945,7 +4093,7 @@ class Choice(Obj): decode_path=decode_path, offset=offset, ) - if tag_only: + if tag_only: # pragma: no cover return value, tail = spec.decode( tlv, @@ -3953,6 +4101,7 @@ class Choice(Obj): leavemm=True, decode_path=sub_decode_path, ctx=ctx, + _ctx_immutable=False, ) obj = self.__class__( schema=self.specs, @@ -3972,6 +4121,7 @@ class Choice(Obj): def pps(self, decode_path=()): yield _pp( + obj=self, asn1_type_name=self.asn1_type_name, obj_name=self.__class__.__name__, decode_path=decode_path, @@ -3985,6 +4135,7 @@ class Choice(Obj): llen=self.llen, vlen=self.vlen, expl_lenindef=self.expl_lenindef, + bered=self.bered, ) if self.ready: yield self.value.pps(decode_path=decode_path + (self.choice,)) @@ -4061,18 +4212,26 @@ class Any(Obj): self.defined = None def _value_sanitize(self, value): + if isinstance(value, binary_type): + return value if isinstance(value, self.__class__): return value._value if isinstance(value, Obj): return value.encode() - if isinstance(value, binary_type): - return value raise InvalidValueType((self.__class__, Obj, binary_type)) @property def ready(self): return self._value is not None + @property + def bered(self): + if self.expl_lenindef or self.lenindef: + return True + if self.defined is None: + return False + return self.defined[1].bered + def copy(self): obj = self.__class__() obj._value = self._value @@ -4082,6 +4241,9 @@ class Any(Obj): obj.offset = self.offset obj.llen = self.llen obj.vlen = self.vlen + obj.expl_lenindef = self.expl_lenindef + obj.lenindef = self.lenindef + obj.ber_encoded = self.ber_encoded return obj def __eq__(self, their): @@ -4145,6 +4307,7 @@ class Any(Obj): decode_path=decode_path + (str(chunk_i),), leavemm=True, ctx=ctx, + _ctx_immutable=False, ) vlen += chunk.tlvlen sub_offset += chunk.tlvlen @@ -4189,6 +4352,7 @@ class Any(Obj): def pps(self, decode_path=()): yield _pp( + obj=self, asn1_type_name=self.asn1_type_name, obj_name=self.__class__.__name__, decode_path=decode_path, @@ -4207,6 +4371,7 @@ class Any(Obj): expl_vlen=self.expl_vlen if self.expled else None, expl_lenindef=self.expl_lenindef, lenindef=self.lenindef, + bered=self.bered, ) defined_by, defined = self.defined or (None, None) if defined_by is not None: @@ -4237,8 +4402,7 @@ def get_def_by_path(defines_by_path, sub_decode_path): def abs_decode_path(decode_path, rel_path): """Create an absolute decode path from current and relative ones - :param decode_path: current decode path, starting point. - Tuple of strings + :param decode_path: current decode path, starting point. Tuple of strings :param rel_path: relative path to ``decode_path``. Tuple of strings. If first tuple's element is "/", then treat it as an absolute path, ignoring ``decode_path`` as @@ -4393,7 +4557,7 @@ class Sequence(Obj): @property def ready(self): - for name, spec in self.specs.items(): + for name, spec in iteritems(self.specs): value = self._value.get(name) if value is None: if spec.optional: @@ -4404,6 +4568,12 @@ class Sequence(Obj): return False return True + @property + def bered(self): + if self.expl_lenindef or self.lenindef or self.ber_encoded: + return True + return any(value.bered for value in itervalues(self._value)) + def copy(self): obj = self.__class__(schema=self.specs) obj.tag = self.tag @@ -4413,7 +4583,10 @@ class Sequence(Obj): obj.offset = self.offset obj.llen = self.llen obj.vlen = self.vlen - obj._value = {k: v.copy() for k, v in self._value.items()} + obj.expl_lenindef = self.expl_lenindef + obj.lenindef = self.lenindef + obj.ber_encoded = self.ber_encoded + obj._value = {k: v.copy() for k, v in iteritems(self._value)} return obj def __eq__(self, their): @@ -4474,7 +4647,7 @@ class Sequence(Obj): def _encoded_values(self): raws = [] - for name, spec in self.specs.items(): + for name, spec in iteritems(self.specs): value = self._value.get(name) if value is None: if spec.optional: @@ -4503,7 +4676,7 @@ class Sequence(Obj): decode_path=decode_path, offset=offset, ) - if tag_only: + if tag_only: # pragma: no cover return lenindef = False ctx_bered = ctx.get("bered", False) @@ -4538,9 +4711,9 @@ class Sequence(Obj): vlen = 0 sub_offset = offset + tlen + llen values = {} - bered = False + ber_encoded = False ctx_allow_default_values = ctx.get("allow_default_values", False) - for name, spec in self.specs.items(): + for name, spec in iteritems(self.specs): if spec.optional and ( (lenindef and v[:EOC_LEN].tobytes() == EOC) or len(v) == 0 @@ -4554,6 +4727,7 @@ class Sequence(Obj): leavemm=True, decode_path=sub_decode_path, ctx=ctx, + _ctx_immutable=False, ) except TagMismatch: if spec.optional: @@ -4578,6 +4752,7 @@ class Sequence(Obj): leavemm=True, decode_path=sub_sub_decode_path, ctx=ctx, + _ctx_immutable=False, ) if len(defined_tail) > 0: raise DecodeError( @@ -4597,6 +4772,7 @@ class Sequence(Obj): leavemm=True, decode_path=sub_decode_path + (DecodePathDefBy(defined_by),), ctx=ctx, + _ctx_immutable=False, ) if len(defined_tail) > 0: raise DecodeError( @@ -4613,7 +4789,7 @@ class Sequence(Obj): v = v_tail if spec.default is not None and value == spec.default: if ctx_bered or ctx_allow_default_values: - bered = True + ber_encoded = True else: raise DecodeError( "DEFAULT value met", @@ -4663,7 +4839,7 @@ class Sequence(Obj): ) obj._value = values obj.lenindef = lenindef - obj.bered = bered + obj.ber_encoded = ber_encoded return obj, tail def __repr__(self): @@ -4678,6 +4854,7 @@ class Sequence(Obj): def pps(self, decode_path=()): yield _pp( + obj=self, asn1_type_name=self.asn1_type_name, obj_name=self.__class__.__name__, decode_path=decode_path, @@ -4695,6 +4872,8 @@ class Sequence(Obj): expl_vlen=self.expl_vlen if self.expled else None, expl_lenindef=self.expl_lenindef, lenindef=self.lenindef, + ber_encoded=self.ber_encoded, + bered=self.bered, ) for name in self.specs: value = self._value.get(name) @@ -4728,6 +4907,9 @@ class Set(Sequence): v = b"".join(raws) return b"".join((self.tag, len_encode(len(v)), v)) + def _specs_items(self): + return iteritems(self.specs) + def _decode(self, tlv, offset, decode_path, ctx, tag_only): try: t, tlen, lv = tag_strip(tlv) @@ -4778,15 +4960,15 @@ class Set(Sequence): vlen = 0 sub_offset = offset + tlen + llen values = {} - bered = False + ber_encoded = False ctx_allow_default_values = ctx.get("allow_default_values", False) ctx_allow_unordered_set = ctx.get("allow_unordered_set", False) value_prev = memoryview(v[:0]) - specs_items = self.specs.items + while len(v) > 0: if lenindef and v[:EOC_LEN].tobytes() == EOC: break - for name, spec in specs_items(): + for name, spec in self._specs_items(): sub_decode_path = decode_path + (name,) try: spec.decode( @@ -4796,6 +4978,7 @@ class Set(Sequence): decode_path=sub_decode_path, ctx=ctx, tag_only=True, + _ctx_immutable=False, ) except TagMismatch: continue @@ -4812,11 +4995,12 @@ class Set(Sequence): leavemm=True, decode_path=sub_decode_path, ctx=ctx, + _ctx_immutable=False, ) value_len = value.fulllen if value_prev.tobytes() > v[:value_len].tobytes(): if ctx_bered or ctx_allow_unordered_set: - bered = True + ber_encoded = True else: raise DecodeError( "unordered " + self.asn1_type_name, @@ -4827,7 +5011,7 @@ class Set(Sequence): if spec.default is None or value != spec.default: pass elif ctx_bered or ctx_allow_default_values: - bered = True + ber_encoded = True else: raise DecodeError( "DEFAULT value met", @@ -4866,7 +5050,7 @@ class Set(Sequence): decode_path=decode_path, offset=offset, ) - obj.bered = bered + obj.ber_encoded = ber_encoded return obj, tail @@ -4965,6 +5149,12 @@ class SequenceOf(Obj): def ready(self): return all(v.ready for v in self._value) + @property + def bered(self): + if self.expl_lenindef or self.lenindef or self.ber_encoded: + return True + return any(v.bered for v in self._value) + def copy(self): obj = self.__class__(schema=self.spec) obj._bound_min = self._bound_min @@ -4976,6 +5166,9 @@ class SequenceOf(Obj): obj.offset = self.offset obj.llen = self.llen obj.vlen = self.vlen + obj.expl_lenindef = self.expl_lenindef + obj.lenindef = self.lenindef + obj.ber_encoded = self.ber_encoded obj._value = [v.copy() for v in self._value] return obj @@ -5050,7 +5243,7 @@ class SequenceOf(Obj): v = b"".join(self._encoded_values()) return b"".join((self.tag, len_encode(len(v)), v)) - def _decode(self, tlv, offset, decode_path, ctx, tag_only): + def _decode(self, tlv, offset, decode_path, ctx, tag_only, ordering_check=False): try: t, tlen, lv = tag_strip(tlv) except DecodeError as err: @@ -5069,10 +5262,11 @@ class SequenceOf(Obj): if tag_only: return lenindef = False + ctx_bered = ctx.get("bered", False) try: l, llen, v = len_decode(lv) except LenIndefForm as err: - if not ctx.get("bered", False): + if not ctx_bered: raise err.__class__( msg=err.msg, klass=self.__class__, @@ -5100,22 +5294,39 @@ class SequenceOf(Obj): vlen = 0 sub_offset = offset + tlen + llen _value = [] + ctx_allow_unordered_set = ctx.get("allow_unordered_set", False) + value_prev = memoryview(v[:0]) + ber_encoded = False spec = self.spec while len(v) > 0: if lenindef and v[:EOC_LEN].tobytes() == EOC: break + sub_decode_path = decode_path + (str(len(_value)),) value, v_tail = spec.decode( v, sub_offset, leavemm=True, - decode_path=decode_path + (str(len(_value)),), + decode_path=sub_decode_path, ctx=ctx, + _ctx_immutable=False, ) value_len = value.fulllen + if ordering_check: + if value_prev.tobytes() > v[:value_len].tobytes(): + if ctx_bered or ctx_allow_unordered_set: + ber_encoded = True + else: + raise DecodeError( + "unordered " + self.asn1_type_name, + klass=self.__class__, + decode_path=sub_decode_path, + offset=sub_offset, + ) + value_prev = v[:value_len] + _value.append(value) sub_offset += value_len vlen += value_len v = v_tail - _value.append(value) try: obj = self.__class__( value=_value, @@ -5144,6 +5355,7 @@ class SequenceOf(Obj): ) obj.lenindef = True tail = v[EOC_LEN:] + obj.ber_encoded = ber_encoded return obj, tail def __repr__(self): @@ -5154,6 +5366,7 @@ class SequenceOf(Obj): def pps(self, decode_path=()): yield _pp( + obj=self, asn1_type_name=self.asn1_type_name, obj_name=self.__class__.__name__, decode_path=decode_path, @@ -5171,6 +5384,8 @@ class SequenceOf(Obj): expl_vlen=self.expl_vlen if self.expled else None, expl_lenindef=self.expl_lenindef, lenindef=self.lenindef, + ber_encoded=self.ber_encoded, + bered=self.bered, ) for i, value in enumerate(self._value): yield value.pps(decode_path=decode_path + (str(i),)) @@ -5193,6 +5408,16 @@ class SetOf(SequenceOf): v = b"".join(raws) return b"".join((self.tag, len_encode(len(v)), v)) + def _decode(self, tlv, offset, decode_path, ctx, tag_only): + return super(SetOf, self)._decode( + tlv, + offset, + decode_path, + ctx, + tag_only, + ordering_check=True, + ) + def obj_by_path(pypath): # pragma: no cover """Import object specified as string Python path @@ -5217,7 +5442,7 @@ def generic_decoder(): # pragma: no cover choice = PrimitiveTypes() choice.specs["SequenceOf"] = SequenceOf(schema=choice) choice.specs["SetOf"] = SetOf(schema=choice) - for i in range(31): + for i in six_xrange(31): choice.specs["SequenceOf%d" % i] = SequenceOf( schema=choice, expl=tag_ctxc(i),