X-Git-Url: http://www.git.cypherpunks.ru/?a=blobdiff_plain;f=pyderasn.py;h=4e9def7bff226e6a53c6b4b4abba12ad3787480b;hb=40b9fcc9381c4456152deb9a448b4e5fd014b7e7;hp=3b7d6cc75dcd5398cd5d440ee742fa26c9a4986c;hpb=5da3ff1e270f6ca27d15d19a7249c0791f46b2a2;p=pyderasn.git diff --git a/pyderasn.py b/pyderasn.py index 3b7d6cc..4e9def7 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 @@ -213,9 +213,11 @@ decoding process. Currently available context options: +* :ref:`allow_default_values ` +* :ref:`allow_expl_oob ` +* :ref:`allow_unordered_set ` * :ref:`bered ` * :ref:`defines_by_path ` -* :ref:`strict_default_existence ` .. _pprinting: @@ -380,19 +382,41 @@ 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`` can contain it. + STRING``, ``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. +.. _allow_expl_oob_ctx: + +Allow explicit tag out-of-bound +------------------------------- + +Invalid BER encoding could contain ``EXPLICIT`` tag containing more than +one value, more than one object. If you set ``allow_expl_oob`` context +option to True, then no error will be raised and that invalid encoding +will be silently further processed. But pay attention that offsets and +lengths will be invalid in that case. + +.. warning:: + + This option should be used only for skipping some decode errors, just + to see the decoded structure somehow. + Primitive types --------------- @@ -491,6 +515,7 @@ Various ------- .. autofunction:: pyderasn.abs_decode_path +.. autofunction:: pyderasn.colonize_hex .. autofunction:: pyderasn.hexenc .. autofunction:: pyderasn.hexdec .. autofunction:: pyderasn.tag_encode @@ -515,9 +540,11 @@ 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 @@ -530,12 +557,13 @@ from six import iterbytes 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 @@ -617,7 +645,11 @@ 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 @@ -672,7 +704,7 @@ class InvalidOID(DecodeError): pass -class ObjUnknown(ValueError): +class ObjUnknown(ASN1Error): def __init__(self, name): super(ObjUnknown, self).__init__() self.name = name @@ -684,7 +716,7 @@ 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 @@ -696,7 +728,7 @@ 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 @@ -710,7 +742,7 @@ 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 @@ -904,7 +936,7 @@ class Obj(object): "vlen", "expl_lenindef", "lenindef", - "bered", + "ber_encoded", ) def __init__( @@ -926,7 +958,7 @@ 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 @@ -938,6 +970,12 @@ 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? @@ -992,6 +1030,7 @@ class Obj(object): decode_path=(), ctx=None, tag_only=False, + _ctx_immutable=True, ): """Decode the data @@ -999,14 +1038,17 @@ 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( @@ -1054,7 +1096,7 @@ 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:] @@ -1089,9 +1131,16 @@ 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): + raise DecodeError( + "explicit tag out-of-bound, longer than data", + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) return obj, (tail if leavemm else tail.tobytes()) @property @@ -1133,7 +1182,10 @@ 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="", @@ -1145,6 +1197,7 @@ class Obj(object): tlen=1, llen=1, vlen=0, + ber_encoded=True, bered=True, ) if self.expl_lenindef: @@ -1156,6 +1209,7 @@ class Obj(object): tlen=1, llen=1, vlen=0, + ber_encoded=True, bered=True, ) @@ -1188,6 +1242,7 @@ class DecodePathDefBy(object): ######################################################################## PP = namedtuple("PP", ( + "obj", "asn1_type_name", "obj_name", "decode_path", @@ -1207,11 +1262,13 @@ PP = namedtuple("PP", ( "expl_vlen", "expl_lenindef", "lenindef", + "ber_encoded", "bered", )) def _pp( + obj=None, asn1_type_name="unknown", obj_name="unknown", decode_path=(), @@ -1231,9 +1288,11 @@ 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, @@ -1253,6 +1312,7 @@ def _pp( expl_vlen, expl_lenindef, lenindef, + ber_encoded, bered, ) @@ -1261,6 +1321,12 @@ 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 range(0, len(hexed), 2)) + + def pp_console_row( pp, oids=None, @@ -1280,7 +1346,9 @@ 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, @@ -1317,7 +1385,7 @@ 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: @@ -1329,6 +1397,15 @@ 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)) @@ -1348,7 +1425,7 @@ 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)) @@ -1356,9 +1433,7 @@ def pp_console_blob(pp, decode_path_len_decrease=0): blob = hexenc(pp.blob).upper() for i in range(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)]) @@ -1580,14 +1655,14 @@ 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", @@ -1603,7 +1678,7 @@ 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): @@ -1611,6 +1686,7 @@ 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, @@ -1628,6 +1704,7 @@ 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): @@ -1937,6 +2014,7 @@ 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, @@ -1954,6 +2032,7 @@ 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 @@ -2268,7 +2347,7 @@ 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: @@ -2279,7 +2358,7 @@ class BitString(Obj): decode_path=decode_path, offset=offset, ) - if tag_only: + if tag_only: # pragma: no cover return lenindef = False try: @@ -2333,6 +2412,7 @@ class BitString(Obj): decode_path=sub_decode_path, leavemm=True, ctx=ctx, + _ctx_immutable=False, ) except TagMismatch: raise DecodeError( @@ -2377,7 +2457,7 @@ 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__, @@ -2397,6 +2477,7 @@ 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, @@ -2416,6 +2497,7 @@ 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) @@ -2696,6 +2778,7 @@ class OctetString(Obj): decode_path=sub_decode_path, leavemm=True, ctx=ctx, + _ctx_immutable=False, ) except TagMismatch: raise DecodeError( @@ -2733,7 +2816,7 @@ 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__, @@ -2746,6 +2829,7 @@ 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, @@ -2765,6 +2849,7 @@ 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) @@ -2859,7 +2944,7 @@ 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) @@ -2890,6 +2975,7 @@ 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, @@ -2905,6 +2991,7 @@ 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 @@ -3109,7 +3196,7 @@ 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) @@ -3141,6 +3228,8 @@ class ObjectIdentifier(Obj): arc = 0 while True: octet = indexbytes(v, i) + if i == 0 and octet == 0x80 and not ctx.get("bered", False): + raise DecodeError("non normalized arc encoding") arc = (arc << 7) | (octet & 0x7F) if octet & 0x80 == 0: arcs.append(arc) @@ -3179,6 +3268,7 @@ 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, @@ -3196,6 +3286,7 @@ 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, + bered=self.bered, ) for pp in self.pps_lenindef(decode_path): yield pp @@ -3404,6 +3495,7 @@ 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, @@ -3421,6 +3513,8 @@ 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 @@ -3433,29 +3527,57 @@ 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 set(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 = set(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 set(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 = set( + (ascii_letters + digits + " '()+,-./:=?").encode("ascii") + ) + + def _value_sanitize(self, value): + value = super(PrintableString, self)._value_sanitize(value) + if not set(value) <= self._allowable_chars: + raise DecodeError("non-printable value") + return value class TeletexString(CommonString): @@ -3606,6 +3728,7 @@ 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, @@ -3623,6 +3746,8 @@ 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 @@ -3821,6 +3946,13 @@ class Choice(Obj): 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 @@ -3910,6 +4042,7 @@ class Choice(Obj): decode_path=sub_decode_path, ctx=ctx, tag_only=True, + _ctx_immutable=False, ) except TagMismatch: continue @@ -3920,7 +4053,7 @@ class Choice(Obj): decode_path=decode_path, offset=offset, ) - if tag_only: + if tag_only: # pragma: no cover return value, tail = spec.decode( tlv, @@ -3928,6 +4061,7 @@ class Choice(Obj): leavemm=True, decode_path=sub_decode_path, ctx=ctx, + _ctx_immutable=False, ) obj = self.__class__( schema=self.specs, @@ -3947,6 +4081,7 @@ 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, @@ -3960,6 +4095,7 @@ 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,)) @@ -4048,6 +4184,14 @@ class Any(Obj): 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 @@ -4120,6 +4264,7 @@ class Any(Obj): decode_path=decode_path + (str(chunk_i),), leavemm=True, ctx=ctx, + _ctx_immutable=False, ) vlen += chunk.tlvlen sub_offset += chunk.tlvlen @@ -4164,6 +4309,7 @@ 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, @@ -4182,6 +4328,7 @@ 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: @@ -4212,8 +4359,7 @@ 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 @@ -4314,18 +4460,14 @@ class Sequence(Obj): All defaulted values are always optional. - .. _strict_default_existence_ctx: + .. _allow_default_values_ctx: - .. warning:: - - When decoded DER contains defaulted value inside, then - technically this is not valid DER encoding. But we allow and pass - it **by default**. Of course reencoding of that kind of DER will - result in different binary representation (validly without - defaulted value inside). You can enable strict defaulted values - existence validation by setting ``"strict_default_existence": - True`` :ref:`context ` option -- decoding process will raise - an exception if defaulted value is met. + DER prohibits default value encoding and will raise an error if + default value is unexpectedly met during decode. + If :ref:`bered ` context option is set, then no error + will be raised, but ``bered`` attribute set. You can disable strict + defaulted values existence validation by setting + ``"allow_default_values": True`` :ref:`context ` option. Two sequences are equal if they have equal specification (schema), implicit/explicit tagging and the same values. @@ -4383,6 +4525,12 @@ 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 self._value.values()) + def copy(self): obj = self.__class__(schema=self.specs) obj.tag = self.tag @@ -4482,13 +4630,14 @@ 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) 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__, @@ -4516,6 +4665,8 @@ class Sequence(Obj): vlen = 0 sub_offset = offset + tlen + llen values = {} + ber_encoded = False + ctx_allow_default_values = ctx.get("allow_default_values", False) for name, spec in self.specs.items(): if spec.optional and ( (lenindef and v[:EOC_LEN].tobytes() == EOC) or @@ -4530,6 +4681,7 @@ class Sequence(Obj): leavemm=True, decode_path=sub_decode_path, ctx=ctx, + _ctx_immutable=False, ) except TagMismatch: if spec.optional: @@ -4554,6 +4706,7 @@ class Sequence(Obj): leavemm=True, decode_path=sub_sub_decode_path, ctx=ctx, + _ctx_immutable=False, ) if len(defined_tail) > 0: raise DecodeError( @@ -4573,6 +4726,7 @@ 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( @@ -4588,15 +4742,15 @@ class Sequence(Obj): sub_offset += value_len v = v_tail if spec.default is not None and value == spec.default: - if ctx.get("strict_default_existence", False): + if ctx_bered or ctx_allow_default_values: + ber_encoded = True + else: raise DecodeError( "DEFAULT value met", klass=self.__class__, decode_path=sub_decode_path, offset=sub_offset, ) - else: - continue values[name] = value spec_defines = getattr(spec, "defines", ()) @@ -4639,6 +4793,7 @@ class Sequence(Obj): ) obj._value = values obj.lenindef = lenindef + obj.ber_encoded = ber_encoded return obj, tail def __repr__(self): @@ -4653,6 +4808,7 @@ 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, @@ -4670,6 +4826,8 @@ 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) @@ -4684,6 +4842,14 @@ class Set(Sequence): """``SET`` structure type Its usage is identical to :py:class:`pyderasn.Sequence`. + + .. _allow_unordered_set_ctx: + + DER prohibits unordered values encoding and will raise an error + during decode. If If :ref:`bered ` context option is set, + then no error will occure. Also you can disable strict values + ordering check by setting ``"allow_unordered_set": True`` + :ref:`context ` option. """ __slots__ = () tag_default = tag_encode(form=TagFormConstructed, num=17) @@ -4714,10 +4880,11 @@ class Set(Sequence): 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__, @@ -4744,6 +4911,10 @@ class Set(Sequence): vlen = 0 sub_offset = offset + tlen + llen values = {} + 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: @@ -4758,6 +4929,7 @@ class Set(Sequence): decode_path=sub_decode_path, ctx=ctx, tag_only=True, + _ctx_immutable=False, ) except TagMismatch: continue @@ -4774,14 +4946,35 @@ 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: + ber_encoded = True + else: + raise DecodeError( + "unordered " + self.asn1_type_name, + klass=self.__class__, + decode_path=sub_decode_path, + offset=sub_offset, + ) + if spec.default is None or value != spec.default: + pass + elif ctx_bered or ctx_allow_default_values: + ber_encoded = True + else: + raise DecodeError( + "DEFAULT value met", + klass=self.__class__, + decode_path=sub_decode_path, + offset=sub_offset, + ) + values[name] = value + value_prev = v[:value_len] 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 - values[name] = value obj = self.__class__( schema=self.specs, impl=self.tag, @@ -4790,7 +4983,6 @@ class Set(Sequence): optional=self.optional, _decoded=(offset, llen, vlen + (EOC_LEN if lenindef else 0)), ) - obj._value = values if lenindef: if v[:EOC_LEN].tobytes() != EOC: raise DecodeError( @@ -4801,6 +4993,7 @@ class Set(Sequence): ) tail = v[EOC_LEN:] obj.lenindef = True + obj._value = values if not obj.ready: raise DecodeError( "not all values are ready", @@ -4808,6 +5001,7 @@ class Set(Sequence): decode_path=decode_path, offset=offset, ) + obj.ber_encoded = ber_encoded return obj, tail @@ -4906,6 +5100,12 @@ 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 @@ -4991,7 +5191,7 @@ 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: @@ -5010,10 +5210,11 @@ 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__, @@ -5041,22 +5242,39 @@ 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, @@ -5085,6 +5303,7 @@ class SequenceOf(Obj): ) obj.lenindef = True tail = v[EOC_LEN:] + obj.ber_encoded = ber_encoded return obj, tail def __repr__(self): @@ -5095,6 +5314,7 @@ 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, @@ -5112,6 +5332,8 @@ 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),)) @@ -5134,6 +5356,16 @@ 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 @@ -5234,18 +5466,23 @@ def main(): # pragma: no cover ) parser.add_argument( "--nobered", - action='store_true', + action="store_true", help="Disallow BER encoding", ) parser.add_argument( "--print-decode-path", - action='store_true', + action="store_true", help="Print decode paths", ) parser.add_argument( "--decode-path-only", help="Print only specified decode path", ) + parser.add_argument( + "--allow-expl-oob", + action="store_true", + help="Allow explicit tag out-of-bound", + ) parser.add_argument( "DERFile", type=argparse.FileType("rb"), @@ -5262,7 +5499,10 @@ def main(): # pragma: no cover pprinter = partial(pprint, big_blobs=True) else: schema, pprinter = generic_decoder() - ctx = {"bered": not args.nobered} + ctx = { + "bered": not args.nobered, + "allow_expl_oob": args.allow_expl_oob, + } 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)