X-Git-Url: http://www.git.cypherpunks.ru/?a=blobdiff_plain;f=pyderasn.py;h=b668507f3a46f9cccdf8d1cfbe3f97ae1cde9826;hb=be964cef343716ff80679f2bc828427bc07c5f02;hp=3cbe67ddd1fbc7386f1df1556f700a54a360c242;hpb=d8f05bb5f06096c6e061c82d40aeb9f43bbb1a21;p=pyderasn.git diff --git a/pyderasn.py b/pyderasn.py index 3cbe67d..b668507 100755 --- a/pyderasn.py +++ b/pyderasn.py @@ -1,6 +1,6 @@ #!/usr/bin/env python # coding: utf-8 -# PyDERASN -- Python ASN.1 DER codec with abstract structures +# PyDERASN -- Python ASN.1 DER/BER codec with abstract structures # Copyright (C) 2017-2018 Sergey Matveev # # This program is free software: you can redistribute it and/or modify @@ -16,10 +16,10 @@ # You should have received a copy of the GNU Lesser General Public # License along with this program. If not, see # . -"""Python ASN.1 DER codec with abstract structures +"""Python ASN.1 DER/BER codec with abstract structures -This library allows you to marshal and unmarshal various structures in -ASN.1 DER format, like this: +This library allows you to marshal various structures in ASN.1 DER +format, unmarshal them in BER/CER/DER ones. >>> i = Integer(123) >>> raw = i.encode() @@ -193,7 +193,7 @@ explicit tag. If you want to know information about it, then use: lesser than ``offset``), ``expl_tlen``, ``expl_llen``, ``expl_vlen`` (that actually equals to ordinary ``tlvlen``). -When error occurs, then :py:exc:`pyderasn.DecodeError` is raised. +When error occurs, :py:exc:`pyderasn.DecodeError` is raised. .. _ctx: @@ -206,6 +206,7 @@ decoding process. Currently available context options: +* :ref:`bered ` * :ref:`defines_by_path ` * :ref:`strict_default_existence ` @@ -363,6 +364,29 @@ First function is useful for path construction when some automatic decoding is already done. ``any`` means literally any value it meet -- useful for SEQUENCE/SET OF-s. +.. _bered_ctx: + +BER encoding +------------ + +By default PyDERASN accepts only DER encoded data. It always encodes to +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`` + attribute is set to True. Only ``BOOLEAN``, ``BIT STRING``, ``OCTET + STRING`` 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. + +EOC (end-of-contents) token's length is taken in advance in object's +value length. + Primitive types --------------- @@ -404,6 +428,10 @@ CommonString ____________ .. autoclass:: pyderasn.CommonString +NumericString +_____________ +.. autoclass:: pyderasn.NumericString + UTCTime _______ .. autoclass:: pyderasn.UTCTime @@ -464,6 +492,17 @@ Various .. autofunction:: pyderasn.tag_ctxp .. autofunction:: pyderasn.tag_ctxc .. autoclass:: pyderasn.Obj +.. autoclass:: pyderasn.DecodeError + :members: __init__ +.. autoclass:: pyderasn.NotEnoughData +.. autoclass:: pyderasn.LenIndefForm +.. autoclass:: pyderasn.TagMismatch +.. autoclass:: pyderasn.InvalidLength +.. autoclass:: pyderasn.InvalidOID +.. autoclass:: pyderasn.ObjUnknown +.. autoclass:: pyderasn.ObjNotReady +.. autoclass:: pyderasn.InvalidValueType +.. autoclass:: pyderasn.BoundsError """ from codecs import getdecoder @@ -516,6 +555,7 @@ __all__ = ( "InvalidOID", "InvalidValueType", "ISO646String", + "LenIndefForm", "NotEnoughData", "Null", "NumericString", @@ -563,6 +603,7 @@ TagClassReprs = { } EOC = b"\x00\x00" EOC_LEN = len(EOC) +LENINDEF = b"\x80" # length indefinite mark ######################################################################## @@ -608,7 +649,7 @@ class NotEnoughData(DecodeError): pass -class LenIndefiniteForm(DecodeError): +class LenIndefForm(DecodeError): pass @@ -803,6 +844,11 @@ def len_encode(l): def len_decode(data): + """Decode length + + :returns: (decoded length, length's length, remaining data) + :raises LenIndefForm: if indefinite form encoding is met + """ if len(data) == 0: raise NotEnoughData("no data at all") first_octet = byte2int(data) @@ -812,7 +858,7 @@ def len_decode(data): if octets_num + 1 > len(data): raise NotEnoughData("encoded length is longer than data") if octets_num == 0: - raise LenIndefiniteForm() + raise LenIndefForm() if byte2int(data[1:]) == 0: raise DecodeError("leading zeros") l = 0 @@ -849,8 +895,9 @@ class Obj(object): "offset", "llen", "vlen", + "expl_lenindef", + "lenindef", "bered", - "expl_bered", ) def __init__( @@ -864,16 +911,15 @@ class Obj(object): self.tag = getattr(self, "impl", self.tag_default) if impl is None else impl self._expl = getattr(self, "expl", None) if expl is None else expl if self.tag != self.tag_default and self._expl is not None: - raise ValueError( - "implicit and explicit tags can not be set simultaneously" - ) + raise ValueError("implicit and explicit tags can not be set simultaneously") if default is not None: optional = True self.optional = optional self.offset, self.llen, self.vlen = _decoded self.default = None + self.expl_lenindef = False + self.lenindef = False self.bered = False - self.expl_bered = False @property def ready(self): # pragma: no cover @@ -984,7 +1030,7 @@ class Obj(object): ) try: l, llen, v = len_decode(lv) - except LenIndefiniteForm as err: + except LenIndefForm as err: if not ctx.get("bered", False): raise err.__class__( msg=err.msg, @@ -1007,12 +1053,13 @@ class Obj(object): eoc_expected, tail = tail[:EOC_LEN], tail[EOC_LEN:] if eoc_expected.tobytes() != EOC: raise DecodeError( - msg="no EOC", + "no EOC", + klass=self.__class__, decode_path=decode_path, offset=offset, ) obj.vlen += EOC_LEN - obj.expl_bered = True + obj.expl_lenindef = True except DecodeError as err: raise err.__class__( msg=err.msg, @@ -1054,7 +1101,7 @@ class Obj(object): @property def expl_llen(self): - if self.expl_bered: + if self.expl_lenindef: return 1 return len(len_encode(self.tlvlen)) @@ -1116,6 +1163,9 @@ PP = namedtuple("PP", ( "expl_tlen", "expl_llen", "expl_vlen", + "expl_lenindef", + "lenindef", + "bered", )) @@ -1137,6 +1187,9 @@ def _pp( expl_tlen=None, expl_llen=None, expl_vlen=None, + expl_lenindef=False, + lenindef=False, + bered=False, ): return PP( asn1_type_name, @@ -1156,6 +1209,9 @@ def _pp( expl_tlen, expl_llen, expl_vlen, + expl_lenindef, + lenindef, + bered, ) @@ -1181,7 +1237,17 @@ def pp_console_row( ) cols.append(_colorize(col, "red", with_colours, ())) col = "[%d,%d,%4d]" % (pp.tlen, pp.llen, pp.vlen) - cols.append(_colorize(col, "green", with_colours, ())) + col = _colorize(col, "green", with_colours, ()) + ber_deoffset = 0 + if pp.expl_lenindef: + ber_deoffset += 2 + if pp.lenindef: + ber_deoffset += 2 + col += ( + " " if ber_deoffset == 0 else + _colorize(("-%d" % ber_deoffset), "red", with_colours) + ) + cols.append(col) if len(pp.decode_path) > 0: cols.append(" ." * (len(pp.decode_path))) ent = pp.decode_path[-1] @@ -1209,6 +1275,8 @@ def pp_console_row( cols.append(_colorize(col, "blue", with_colours)) if pp.asn1_type_name.replace(" ", "") != pp.obj_name.upper(): cols.append(_colorize(pp.obj_name, "magenta", with_colours)) + if pp.bered: + cols.append(_colorize("BER", "red", with_colours)) cols.append(_colorize(pp.asn1_type_name, "cyan", with_colours)) if pp.value is not None: value = pp.value @@ -1232,7 +1300,7 @@ def pp_console_row( def pp_console_blob(pp): - cols = [" " * len("XXXXXYY [X,X,XXXX]")] + cols = [" " * len("XXXXXYY [X,X,XXXX]YY")] if len(pp.decode_path) > 0: cols.append(" ." * (len(pp.decode_path) + 1)) if isinstance(pp.blob, binary_type): @@ -1487,6 +1555,8 @@ class Boolean(Obj): expl_tlen=self.expl_tlen if self.expled else None, 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, ) @@ -1809,6 +1879,7 @@ class Integer(Obj): expl_tlen=self.expl_tlen if self.expled else None, 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, ) @@ -1824,6 +1895,8 @@ class BitString(Obj): >>> b.bit_len 88 + >>> BitString("'0A3B5F291CD'H") + BIT STRING 44 bits 0a3b5f291cd0 >>> b = BitString("'010110000000'B") BIT STRING 12 bits 5800 >>> b.bit_len @@ -1850,6 +1923,14 @@ class BitString(Obj): ['nonRepudiation', 'keyEncipherment'] >>> b.specs {'nonRepudiation': 1, 'digitalSignature': 0, 'keyEncipherment': 2} + + .. note:: + + Pay attention that BIT STRING can be encoded both in primitive + and constructed forms. Decoder always checks constructed form tag + additionally to specified primitive one. If BER decoding is + :ref:`not enabled `, then decoder will fail, because + of DER restrictions. """ __slots__ = ("tag_constructed", "specs", "defined") tag_default = tag_encode(3) @@ -1912,21 +1993,23 @@ class BitString(Obj): if isinstance(value, (string_types, binary_type)): if ( isinstance(value, string_types) and - value.startswith("'") and - value.endswith("'B") + value.startswith("'") ): - value = value[1:-2] - if not set(value) <= set(("0", "1")): - raise ValueError("B's coding contains unacceptable chars") - return self._bits2octets(value) - elif isinstance(value, binary_type): + if value.endswith("'B"): + value = value[1:-2] + if not set(value) <= set(("0", "1")): + raise ValueError("B's coding contains unacceptable chars") + return self._bits2octets(value) + elif value.endswith("'H"): + value = value[1:-2] + return ( + len(value) * 4, + hexdec(value + ("" if len(value) % 2 == 0 else "0")), + ) + if isinstance(value, binary_type): return (len(value) * 8, value) else: - raise InvalidValueType(( - self.__class__, - string_types, - binary_type, - )) + raise InvalidValueType((self.__class__, string_types, binary_type)) if isinstance(value, tuple): if ( len(value) == 2 and @@ -2115,18 +2198,19 @@ class BitString(Obj): if t == self.tag_constructed: if not ctx.get("bered", False): raise DecodeError( - msg="unallowed BER constructed encoding", + "unallowed BER constructed encoding", + klass=self.__class__, decode_path=decode_path, offset=offset, ) if tag_only: return - eoc_expected = False + lenindef = False try: l, llen, v = len_decode(lv) - except LenIndefiniteForm: + except LenIndefForm: llen, l, v = 1, 0, lv[1:] - eoc_expected = True + lenindef = True except DecodeError as err: raise err.__class__( msg=err.msg, @@ -2134,14 +2218,14 @@ class BitString(Obj): decode_path=decode_path, offset=offset, ) - if l > 0 and l > len(v): + if l > len(v): raise NotEnoughData( "encoded length is longer than data", klass=self.__class__, decode_path=decode_path, offset=offset, ) - if not eoc_expected and l == 0: + if not lenindef and l == 0: raise NotEnoughData( "zero length", klass=self.__class__, @@ -2152,7 +2236,7 @@ class BitString(Obj): sub_offset = offset + tlen + llen vlen = 0 while True: - if eoc_expected: + if lenindef: if v[:EOC_LEN].tobytes() == EOC: break else: @@ -2160,8 +2244,9 @@ class BitString(Obj): break if vlen > l: raise DecodeError( - msg="chunk out of bounds", - decode_path=len(chunks) - 1, + "chunk out of bounds", + klass=self.__class__, + decode_path=decode_path + (str(len(chunks) - 1),), offset=chunks[-1].offset, ) sub_decode_path = decode_path + (str(len(chunks)),) @@ -2175,7 +2260,8 @@ class BitString(Obj): ) except TagMismatch: raise DecodeError( - msg="expected BitString encoded chunk", + "expected BitString encoded chunk", + klass=self.__class__, decode_path=sub_decode_path, offset=sub_offset, ) @@ -2185,7 +2271,8 @@ class BitString(Obj): v = v_tail if len(chunks) == 0: raise DecodeError( - msg="no chunks", + "no chunks", + klass=self.__class__, decode_path=decode_path, offset=offset, ) @@ -2194,7 +2281,8 @@ class BitString(Obj): for chunk_i, chunk in enumerate(chunks[:-1]): if chunk.bit_len % 8 != 0: raise DecodeError( - msg="BitString chunk is not multiple of 8 bit", + "BitString chunk is not multiple of 8 bits", + klass=self.__class__, decode_path=decode_path + (str(chunk_i),), offset=chunk.offset, ) @@ -2210,10 +2298,11 @@ class BitString(Obj): default=self.default, optional=self.optional, _specs=self.specs, - _decoded=(offset, llen, vlen + (EOC_LEN if eoc_expected else 0)), + _decoded=(offset, llen, vlen + (EOC_LEN if lenindef else 0)), ) + obj.lenindef = lenindef obj.bered = True - return obj, v[EOC_LEN if eoc_expected else 0:] + return obj, (v[EOC_LEN:] if lenindef else v) raise TagMismatch( klass=self.__class__, decode_path=decode_path, @@ -2249,6 +2338,9 @@ class BitString(Obj): expl_tlen=self.expl_tlen if self.expled else None, 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, + lenindef=self.lenindef, + bered=self.bered, ) defined_by, defined = self.defined or (None, None) if defined_by is not None: @@ -2272,6 +2364,14 @@ class OctetString(Obj): pyderasn.BoundsError: unsatisfied bounds: 4 <= 5 <= 4 >>> OctetString(b"hell", bounds=(4, 4)) OCTET STRING 4 bytes 68656c6c + + .. note:: + + Pay attention that OCTET STRING can be encoded both in primitive + and constructed forms. Decoder always checks constructed form tag + additionally to specified primitive one. If BER decoding is + :ref:`not enabled `, then decoder will fail, because + of DER restrictions. """ __slots__ = ("tag_constructed", "_bound_min", "_bound_max", "defined") tag_default = tag_encode(4) @@ -2466,18 +2566,19 @@ class OctetString(Obj): if t == self.tag_constructed: if not ctx.get("bered", False): raise DecodeError( - msg="unallowed BER constructed encoding", + "unallowed BER constructed encoding", + klass=self.__class__, decode_path=decode_path, offset=offset, ) if tag_only: return - eoc_expected = False + lenindef = False try: l, llen, v = len_decode(lv) - except LenIndefiniteForm: + except LenIndefForm: llen, l, v = 1, 0, lv[1:] - eoc_expected = True + lenindef = True except DecodeError as err: raise err.__class__( msg=err.msg, @@ -2485,25 +2586,18 @@ class OctetString(Obj): decode_path=decode_path, offset=offset, ) - if l > 0 and l > len(v): + if l > len(v): raise NotEnoughData( "encoded length is longer than data", klass=self.__class__, decode_path=decode_path, offset=offset, ) - if not eoc_expected and l == 0: - raise NotEnoughData( - "zero length", - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) chunks = [] sub_offset = offset + tlen + llen vlen = 0 while True: - if eoc_expected: + if lenindef: if v[:EOC_LEN].tobytes() == EOC: break else: @@ -2511,8 +2605,9 @@ class OctetString(Obj): break if vlen > l: raise DecodeError( - msg="chunk out of bounds", - decode_path=len(chunks) - 1, + "chunk out of bounds", + klass=self.__class__, + decode_path=decode_path + (str(len(chunks) - 1),), offset=chunks[-1].offset, ) sub_decode_path = decode_path + (str(len(chunks)),) @@ -2526,7 +2621,8 @@ class OctetString(Obj): ) except TagMismatch: raise DecodeError( - msg="expected OctetString encoded chunk", + "expected OctetString encoded chunk", + klass=self.__class__, decode_path=sub_decode_path, offset=sub_offset, ) @@ -2534,12 +2630,6 @@ class OctetString(Obj): sub_offset += chunk.tlvlen vlen += chunk.tlvlen v = v_tail - if len(chunks) == 0: - raise DecodeError( - msg="no chunks", - decode_path=decode_path, - offset=offset, - ) try: obj = self.__class__( value=b"".join(bytes(chunk) for chunk in chunks), @@ -2548,7 +2638,7 @@ class OctetString(Obj): expl=self._expl, default=self.default, optional=self.optional, - _decoded=(offset, llen, vlen + (EOC_LEN if eoc_expected else 0)), + _decoded=(offset, llen, vlen + (EOC_LEN if lenindef else 0)), ) except DecodeError as err: raise DecodeError( @@ -2564,8 +2654,9 @@ class OctetString(Obj): decode_path=decode_path, offset=offset, ) + obj.lenindef = lenindef obj.bered = True - return obj, v[EOC_LEN if eoc_expected else 0:] + return obj, (v[EOC_LEN:] if lenindef else v) raise TagMismatch( klass=self.__class__, decode_path=decode_path, @@ -2594,6 +2685,9 @@ class OctetString(Obj): expl_tlen=self.expl_tlen if self.expled else None, 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, + lenindef=self.lenindef, + bered=self.bered, ) defined_by, defined = self.defined or (None, None) if defined_by is not None: @@ -2730,6 +2824,7 @@ class Null(Obj): expl_tlen=self.expl_tlen if self.expled else None, 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, ) @@ -3018,6 +3113,7 @@ class ObjectIdentifier(Obj): expl_tlen=self.expl_tlen if self.expled else None, 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, ) @@ -3236,6 +3332,11 @@ class CommonString(OctetString): tlen=self.tlen, llen=self.llen, vlen=self.vlen, + expl_offset=self.expl_offset if self.expled else None, + expl_tlen=self.expl_tlen if self.expled else None, + 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, ) @@ -3247,6 +3348,10 @@ class UTF8String(CommonString): class NumericString(CommonString): + """Numeric string + + Its value is properly sanitized: only ASCII digits can be stored. + """ __slots__ = () tag_default = tag_encode(18) encoding = "ascii" @@ -3424,6 +3529,11 @@ class UTCTime(CommonString): tlen=self.tlen, llen=self.llen, vlen=self.vlen, + expl_offset=self.expl_offset if self.expled else None, + expl_tlen=self.expl_tlen if self.expled else None, + 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, ) @@ -3755,6 +3865,7 @@ class Choice(Obj): tlen=self.tlen, llen=self.llen, vlen=self.vlen, + expl_lenindef=self.expl_lenindef, ) if self.ready: yield self.value.pps(decode_path=decode_path + (self.choice,)) @@ -3886,7 +3997,47 @@ class Any(Obj): def _decode(self, tlv, offset, decode_path, ctx, tag_only): try: t, tlen, lv = tag_strip(tlv) + except DecodeError as err: + raise err.__class__( + msg=err.msg, + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + try: l, llen, v = len_decode(lv) + except LenIndefForm as err: + if not ctx.get("bered", False): + raise err.__class__( + msg=err.msg, + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + llen, vlen, v = 1, 0, lv[1:] + sub_offset = offset + tlen + llen + chunk_i = 0 + while v[:EOC_LEN].tobytes() != EOC: + chunk, v = Any().decode( + v, + offset=sub_offset, + decode_path=decode_path + (str(chunk_i),), + leavemm=True, + ctx=ctx, + ) + vlen += chunk.tlvlen + sub_offset += chunk.tlvlen + chunk_i += 1 + tlvlen = tlen + llen + vlen + EOC_LEN + obj = self.__class__( + value=tlv[:tlvlen].tobytes(), + expl=self._expl, + optional=self.optional, + _decoded=(offset, 0, tlvlen), + ) + obj.lenindef = True + obj.tag = t + return obj, v[EOC_LEN:] except DecodeError as err: raise err.__class__( msg=err.msg, @@ -3933,6 +4084,8 @@ class Any(Obj): expl_tlen=self.expl_tlen if self.expled else None, 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, + lenindef=self.lenindef, ) defined_by, defined = self.defined or (None, None) if defined_by is not None: @@ -4233,8 +4386,19 @@ class Sequence(Obj): ) if tag_only: return + lenindef = False try: l, llen, v = len_decode(lv) + except LenIndefForm as err: + if not ctx.get("bered", False): + raise err.__class__( + msg=err.msg, + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + l, llen, v = 0, 1, lv[1:] + lenindef = True except DecodeError as err: raise err.__class__( msg=err.msg, @@ -4249,11 +4413,16 @@ class Sequence(Obj): decode_path=decode_path, offset=offset, ) - v, tail = v[:l], v[l:] + if not lenindef: + v, tail = v[:l], v[l:] + vlen = 0 sub_offset = offset + tlen + llen values = {} for name, spec in self.specs.items(): - if len(v) == 0 and spec.optional: + if spec.optional and ( + (lenindef and v[:EOC_LEN].tobytes() == EOC) or + len(v) == 0 + ): continue sub_decode_path = decode_path + (name,) try: @@ -4316,7 +4485,9 @@ class Sequence(Obj): ) value.defined = (defined_by, defined_value) - sub_offset += (value.expl_tlvlen if value.expled else value.tlvlen) + value_len = value.expl_tlvlen if value.expled else value.tlvlen + vlen += value_len + sub_offset += value_len v = v_tail if spec.default is not None and value == spec.default: if ctx.get("strict_default_existence", False): @@ -4343,7 +4514,17 @@ class Sequence(Obj): abs_decode_path(sub_decode_path[:-1], rel_path), (value, defined), )) - if len(v) > 0: + if lenindef: + if v[:EOC_LEN].tobytes() != EOC: + raise DecodeError( + "no EOC", + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + tail = v[EOC_LEN:] + vlen += EOC_LEN + elif len(v) > 0: raise DecodeError( "remaining data", klass=self.__class__, @@ -4356,9 +4537,10 @@ class Sequence(Obj): expl=self._expl, default=self.default, optional=self.optional, - _decoded=(offset, llen, l), + _decoded=(offset, llen, vlen), ) obj._value = values + obj.lenindef = lenindef return obj, tail def __repr__(self): @@ -4388,6 +4570,8 @@ class Sequence(Obj): expl_tlen=self.expl_tlen if self.expled else None, 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, + lenindef=self.lenindef, ) for name in self.specs: value = self._value.get(name) @@ -4429,8 +4613,19 @@ class Set(Sequence): ) if tag_only: return + lenindef = False try: l, llen, v = len_decode(lv) + except LenIndefForm as err: + if not ctx.get("bered", False): + raise err.__class__( + msg=err.msg, + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + l, llen, v = 0, 1, lv[1:] + lenindef = True except DecodeError as err: raise err.__class__( msg=err.msg, @@ -4444,11 +4639,15 @@ class Set(Sequence): klass=self.__class__, offset=offset, ) - v, tail = v[:l], v[l:] + if not lenindef: + v, tail = v[:l], v[l:] + vlen = 0 sub_offset = offset + tlen + llen values = {} specs_items = self.specs.items while len(v) > 0: + if lenindef and v[:EOC_LEN].tobytes() == EOC: + break for name, spec in specs_items(): sub_decode_path = decode_path + (name,) try: @@ -4476,9 +4675,9 @@ class Set(Sequence): decode_path=sub_decode_path, ctx=ctx, ) - sub_offset += ( - value.expl_tlvlen if value.expled else value.tlvlen - ) + value_len = value.expl_tlvlen if value.expled else value.tlvlen + sub_offset += value_len + vlen += value_len v = v_tail if spec.default is None or value != spec.default: # pragma: no cover # SeqMixing.test_encoded_default_accepted covers that place @@ -4489,9 +4688,26 @@ class Set(Sequence): expl=self._expl, default=self.default, optional=self.optional, - _decoded=(offset, llen, l), + _decoded=(offset, llen, vlen + (EOC_LEN if lenindef else 0)), ) obj._value = values + if lenindef: + if v[:EOC_LEN].tobytes() != EOC: + raise DecodeError( + "no EOC", + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + tail = v[EOC_LEN:] + obj.lenindef = True + if not obj.ready: + raise DecodeError( + "not all values are ready", + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) return obj, tail @@ -4693,8 +4909,19 @@ class SequenceOf(Obj): ) if tag_only: return + lenindef = False try: l, llen, v = len_decode(lv) + except LenIndefForm as err: + if not ctx.get("bered", False): + raise err.__class__( + msg=err.msg, + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + l, llen, v = 0, 1, lv[1:] + lenindef = True except DecodeError as err: raise err.__class__( msg=err.msg, @@ -4709,11 +4936,15 @@ class SequenceOf(Obj): decode_path=decode_path, offset=offset, ) - v, tail = v[:l], v[l:] + if not lenindef: + v, tail = v[:l], v[l:] + vlen = 0 sub_offset = offset + tlen + llen _value = [] spec = self.spec while len(v) > 0: + if lenindef and v[:EOC_LEN].tobytes() == EOC: + break value, v_tail = spec.decode( v, sub_offset, @@ -4721,7 +4952,9 @@ class SequenceOf(Obj): decode_path=decode_path + (str(len(_value)),), ctx=ctx, ) - sub_offset += (value.expl_tlvlen if value.expled else value.tlvlen) + value_len = value.expl_tlvlen if value.expled else value.tlvlen + sub_offset += value_len + vlen += value_len v = v_tail _value.append(value) obj = self.__class__( @@ -4732,8 +4965,18 @@ class SequenceOf(Obj): expl=self._expl, default=self.default, optional=self.optional, - _decoded=(offset, llen, l), + _decoded=(offset, llen, vlen + (EOC_LEN if lenindef else 0)), ) + if lenindef: + if v[:EOC_LEN].tobytes() != EOC: + raise DecodeError( + "no EOC", + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + obj.lenindef = True + tail = v[EOC_LEN:] return obj, tail def __repr__(self): @@ -4759,6 +5002,8 @@ class SequenceOf(Obj): expl_tlen=self.expl_tlen if self.expled else None, 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, + lenindef=self.lenindef, ) for i, value in enumerate(self._value): yield value.pps(decode_path=decode_path + (str(i),)) @@ -4842,7 +5087,7 @@ def generic_decoder(): # pragma: no cover def main(): # pragma: no cover import argparse - parser = argparse.ArgumentParser(description="PyDERASN ASN.1 DER decoder") + parser = argparse.ArgumentParser(description="PyDERASN ASN.1 BER/DER decoder") parser.add_argument( "--skip", type=int, @@ -4861,6 +5106,11 @@ def main(): # pragma: no cover "--defines-by-path", help="Python path to decoder's defines_by_path", ) + parser.add_argument( + "--nobered", + action='store_true', + help="Disallow BER encoding", + ) parser.add_argument( "DERFile", type=argparse.FileType("rb"), @@ -4877,13 +5127,10 @@ def main(): # pragma: no cover pprinter = partial(pprint, big_blobs=True) else: schema, pprinter = generic_decoder() - obj, tail = schema().decode( - der, - ctx=( - None if args.defines_by_path is None else - {"defines_by_path": obj_by_path(args.defines_by_path)} - ), - ) + ctx = {"bered": not args.nobered} + if args.defines_by_path is not None: + ctx["defines_by_path"] = obj_by_path(args.defines_by_path) + obj, tail = schema().decode(der, ctx=ctx) print(pprinter( obj, oids=oids,