X-Git-Url: http://www.git.cypherpunks.ru/?a=blobdiff_plain;f=pyderasn.py;h=4142af690d616b316ac46494dfde8857fcb25a02;hb=863d21e01e95ba9f1626159b6997c794e0cc11a5;hp=b9bc9dae515564a24dcc052005754d54993fca03;hpb=a34b3e75db1e13750560c89c2b80821f2e5d0d1f;p=pyderasn.git diff --git a/pyderasn.py b/pyderasn.py index b9bc9da..4142af6 100755 --- a/pyderasn.py +++ b/pyderasn.py @@ -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: @@ -274,7 +276,7 @@ You can specify multiple fields, that will be autodecoded -- that is why ``defines`` kwarg is a sequence. You can specify defined field relatively or absolutely to current decode path. For example ``defines`` for AlgorithmIdentifier of X.509's -``tbsCertificate.subjectPublicKeyInfo.algorithm.algorithm``:: +``tbsCertificate:subjectPublicKeyInfo:algorithm:algorithm``:: ( (("parameters",), { @@ -302,8 +304,7 @@ Following types can be automatically decoded (DEFINED BY): When any of those fields is automatically decoded, then ``.defined`` attribute contains ``(OID, value)`` tuple. ``OID`` tells by which OID it was defined, ``value`` contains corresponding decoded value. For example -above, ``content_info["content"].defined == (id_signedData, -signed_data)``. +above, ``content_info["content"].defined == (id_signedData, signed_data)``. .. _defines_by_path_ctx: @@ -381,19 +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 --------------- @@ -516,9 +539,11 @@ 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 @@ -531,12 +556,13 @@ 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 @@ -618,7 +644,11 @@ 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 @@ -641,7 +671,7 @@ class DecodeError(Exception): c for c in ( "" if self.klass is None else self.klass.__name__, ( - ("(%s)" % ".".join(str(dp) for dp in self.decode_path)) + ("(%s)" % ":".join(str(dp) for dp in self.decode_path)) if len(self.decode_path) > 0 else "" ), ("(at %d)" % self.offset) if self.offset > 0 else "", @@ -673,7 +703,7 @@ class InvalidOID(DecodeError): pass -class ObjUnknown(ValueError): +class ObjUnknown(ASN1Error): def __init__(self, name): super(ObjUnknown, self).__init__() self.name = name @@ -685,7 +715,7 @@ 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 @@ -697,7 +727,7 @@ 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 @@ -711,7 +741,7 @@ 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 @@ -905,7 +935,7 @@ class Obj(object): "vlen", "expl_lenindef", "lenindef", - "bered", + "ber_encoded", ) def __init__( @@ -927,7 +957,7 @@ 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 @@ -939,6 +969,12 @@ 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? @@ -993,6 +1029,7 @@ class Obj(object): decode_path=(), ctx=None, tag_only=False, + _ctx_immutable=True, ): """Decode the data @@ -1000,14 +1037,17 @@ 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( @@ -1055,7 +1095,7 @@ 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:] @@ -1090,9 +1130,16 @@ 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 @@ -1134,7 +1181,10 @@ 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="", @@ -1146,6 +1196,7 @@ class Obj(object): tlen=1, llen=1, vlen=0, + ber_encoded=True, bered=True, ) if self.expl_lenindef: @@ -1157,6 +1208,7 @@ class Obj(object): tlen=1, llen=1, vlen=0, + ber_encoded=True, bered=True, ) @@ -1208,6 +1260,7 @@ PP = namedtuple("PP", ( "expl_vlen", "expl_lenindef", "lenindef", + "ber_encoded", "bered", )) @@ -1232,6 +1285,7 @@ def _pp( expl_vlen=None, expl_lenindef=False, lenindef=False, + ber_encoded=False, bered=False, ): return PP( @@ -1254,11 +1308,12 @@ def _pp( expl_vlen, expl_lenindef, lenindef, + ber_encoded, bered, ) -def _colorize(what, colour, with_colours, attrs=("bold",)): +def _colourize(what, colour, with_colours, attrs=("bold",)): return colored(what, colour, attrs=attrs) if with_colours else what @@ -1268,6 +1323,8 @@ def pp_console_row( with_offsets=False, with_blob=True, with_colours=False, + with_decode_path=False, + decode_path_len_decrease=0, ): cols = [] if with_offsets: @@ -1279,20 +1336,23 @@ def pp_console_row( ), LENINDEF_PP_CHAR if pp.expl_lenindef else " ", ) - cols.append(_colorize(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, pp.vlen, LENINDEF_PP_CHAR if pp.lenindef else " " ) - col = _colorize(col, "green", with_colours, ()) + col = _colourize(col, "green", with_colours, ()) cols.append(col) - if len(pp.decode_path) > 0: - cols.append(" ." * (len(pp.decode_path))) + decode_path_len = len(pp.decode_path) - decode_path_len_decrease + if decode_path_len > 0: + cols.append(" ." * decode_path_len) ent = pp.decode_path[-1] if isinstance(ent, DecodePathDefBy): - cols.append(_colorize("DEFINED BY", "red", with_colours, ("reverse",))) + cols.append(_colourize("DEFINED BY", "red", with_colours, ("reverse",))) value = str(ent.defined_by) if ( oids is not None and @@ -1300,49 +1360,56 @@ def pp_console_row( ObjectIdentifier.asn1_type_name and value in oids ): - cols.append(_colorize("%s:" % oids[value], "green", with_colours)) + cols.append(_colourize("%s:" % oids[value], "green", with_colours)) else: - cols.append(_colorize("%s:" % value, "white", with_colours, ("reverse",))) + cols.append(_colourize("%s:" % value, "white", with_colours, ("reverse",))) else: - cols.append(_colorize("%s:" % ent, "yellow", with_colours, ("reverse",))) + cols.append(_colourize("%s:" % ent, "yellow", with_colours, ("reverse",))) if pp.expl is not None: klass, _, num = pp.expl col = "[%s%d] EXPLICIT" % (TagClassReprs[klass], num) - cols.append(_colorize(col, "blue", with_colours)) + cols.append(_colourize(col, "blue", with_colours)) if pp.impl is not None: klass, _, num = pp.impl col = "[%s%d]" % (TagClassReprs[klass], num) - cols.append(_colorize(col, "blue", with_colours)) + cols.append(_colourize(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)) + cols.append(_colourize(pp.obj_name, "magenta", with_colours)) + 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: value = pp.value - cols.append(_colorize(value, "white", with_colours, ("reverse",))) + cols.append(_colourize(value, "white", with_colours, ("reverse",))) if ( oids is not None and pp.asn1_type_name == ObjectIdentifier.asn1_type_name and value in oids ): - cols.append(_colorize("(%s)" % oids[value], "green", with_colours)) + cols.append(_colourize("(%s)" % oids[value], "green", with_colours)) if with_blob: if isinstance(pp.blob, binary_type): cols.append(hexenc(pp.blob)) elif isinstance(pp.blob, tuple): cols.append(", ".join(pp.blob)) if pp.optional: - cols.append(_colorize("OPTIONAL", "red", with_colours)) + cols.append(_colourize("OPTIONAL", "red", with_colours)) if pp.default: - cols.append(_colorize("DEFAULT", "red", with_colours)) + cols.append(_colourize("DEFAULT", "red", with_colours)) + if with_decode_path: + cols.append(_colourize( + "[%s]" % ":".join(str(p) for p in pp.decode_path), + "grey", + with_colours, + )) return " ".join(cols) -def pp_console_blob(pp): - cols = [" " * len("XXXXXYYZ [X,X,XXXX]Z")] - if len(pp.decode_path) > 0: - cols.append(" ." * (len(pp.decode_path) + 1)) +def pp_console_blob(pp, decode_path_len_decrease=0): + 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): @@ -1354,7 +1421,14 @@ def pp_console_blob(pp): yield " ".join(cols + [", ".join(pp.blob)]) -def pprint(obj, oids=None, big_blobs=False, with_colours=False): +def pprint( + obj, + oids=None, + big_blobs=False, + with_colours=False, + with_decode_path=False, + decode_path_only=(), +): """Pretty print object :param Obj obj: object you want to pretty print @@ -1365,10 +1439,19 @@ def pprint(obj, oids=None, big_blobs=False, with_colours=False): lines :param with_colours: colourize output, if ``termcolor`` library is available + :param with_decode_path: print decode path + :param decode_path_only: print only that specified decode path """ def _pprint_pps(pps): for pp in pps: if hasattr(pp, "_fields"): + if ( + decode_path_only != () and + tuple( + str(p) for p in pp.decode_path[:len(decode_path_only)] + ) != decode_path_only + ): + continue if big_blobs: yield pp_console_row( pp, @@ -1376,8 +1459,13 @@ def pprint(obj, oids=None, big_blobs=False, with_colours=False): with_offsets=True, with_blob=False, with_colours=with_colours, + with_decode_path=with_decode_path, + decode_path_len_decrease=len(decode_path_only), ) - for row in pp_console_blob(pp): + for row in pp_console_blob( + pp, + decode_path_len_decrease=len(decode_path_only), + ): yield row else: yield pp_console_row( @@ -1386,6 +1474,8 @@ def pprint(obj, oids=None, big_blobs=False, with_colours=False): with_offsets=True, with_blob=True, with_colours=with_colours, + with_decode_path=with_decode_path, + decode_path_len_decrease=len(decode_path_only), ) else: for row in _pprint_pps(pp): @@ -1548,14 +1638,14 @@ 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", @@ -1571,7 +1661,7 @@ 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): @@ -1596,6 +1686,7 @@ 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): @@ -1922,6 +2013,7 @@ 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 @@ -2236,7 +2328,7 @@ 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: @@ -2247,7 +2339,7 @@ class BitString(Obj): decode_path=decode_path, offset=offset, ) - if tag_only: + if tag_only: # pragma: no cover return lenindef = False try: @@ -2301,6 +2393,7 @@ class BitString(Obj): decode_path=sub_decode_path, leavemm=True, ctx=ctx, + _ctx_immutable=False, ) except TagMismatch: raise DecodeError( @@ -2345,7 +2438,7 @@ 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__, @@ -2384,6 +2477,7 @@ 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) @@ -2664,6 +2758,7 @@ class OctetString(Obj): decode_path=sub_decode_path, leavemm=True, ctx=ctx, + _ctx_immutable=False, ) except TagMismatch: raise DecodeError( @@ -2701,7 +2796,7 @@ 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__, @@ -2733,6 +2828,7 @@ 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) @@ -2827,7 +2923,7 @@ 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) @@ -2873,6 +2969,7 @@ 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 @@ -3077,7 +3174,7 @@ 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) @@ -3164,6 +3261,7 @@ 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 @@ -3389,6 +3487,8 @@ 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 @@ -3401,29 +3501,57 @@ 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): @@ -3591,6 +3719,8 @@ 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 @@ -3789,6 +3919,13 @@ 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 @@ -3878,6 +4015,7 @@ class Choice(Obj): decode_path=sub_decode_path, ctx=ctx, tag_only=True, + _ctx_immutable=False, ) except TagMismatch: continue @@ -3888,7 +4026,7 @@ class Choice(Obj): decode_path=decode_path, offset=offset, ) - if tag_only: + if tag_only: # pragma: no cover return value, tail = spec.decode( tlv, @@ -3896,6 +4034,7 @@ class Choice(Obj): leavemm=True, decode_path=sub_decode_path, ctx=ctx, + _ctx_immutable=False, ) obj = self.__class__( schema=self.specs, @@ -3928,6 +4067,7 @@ 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,)) @@ -4016,6 +4156,14 @@ 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 @@ -4088,6 +4236,7 @@ class Any(Obj): decode_path=decode_path + (str(chunk_i),), leavemm=True, ctx=ctx, + _ctx_immutable=False, ) vlen += chunk.tlvlen sub_offset += chunk.tlvlen @@ -4150,6 +4299,7 @@ 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: @@ -4180,8 +4330,7 @@ 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 @@ -4282,18 +4431,14 @@ class Sequence(Obj): All defaulted values are always optional. - .. _strict_default_existence_ctx: - - .. warning:: + .. _allow_default_values_ctx: - 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. @@ -4351,6 +4496,12 @@ 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 @@ -4450,13 +4601,14 @@ 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__, @@ -4484,6 +4636,8 @@ 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 @@ -4498,6 +4652,7 @@ class Sequence(Obj): leavemm=True, decode_path=sub_decode_path, ctx=ctx, + _ctx_immutable=False, ) except TagMismatch: if spec.optional: @@ -4522,6 +4677,7 @@ class Sequence(Obj): leavemm=True, decode_path=sub_sub_decode_path, ctx=ctx, + _ctx_immutable=False, ) if len(defined_tail) > 0: raise DecodeError( @@ -4541,6 +4697,7 @@ 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( @@ -4556,15 +4713,15 @@ 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", ()) @@ -4607,6 +4764,7 @@ class Sequence(Obj): ) obj._value = values obj.lenindef = lenindef + obj.ber_encoded = ber_encoded return obj, tail def __repr__(self): @@ -4638,6 +4796,8 @@ 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) @@ -4652,6 +4812,14 @@ 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) @@ -4682,10 +4850,11 @@ 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__, @@ -4712,6 +4881,10 @@ 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: @@ -4726,6 +4899,7 @@ class Set(Sequence): decode_path=sub_decode_path, ctx=ctx, tag_only=True, + _ctx_immutable=False, ) except TagMismatch: continue @@ -4742,14 +4916,35 @@ 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, @@ -4758,7 +4953,6 @@ 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( @@ -4769,6 +4963,7 @@ class Set(Sequence): ) tail = v[EOC_LEN:] obj.lenindef = True + obj._value = values if not obj.ready: raise DecodeError( "not all values are ready", @@ -4776,6 +4971,7 @@ class Set(Sequence): decode_path=decode_path, offset=offset, ) + obj.ber_encoded = ber_encoded return obj, tail @@ -4874,6 +5070,12 @@ 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 @@ -4959,7 +5161,7 @@ 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: @@ -4978,10 +5180,11 @@ 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__, @@ -5009,32 +5212,57 @@ 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) - obj = self.__class__( - value=_value, - schema=spec, - bounds=(self._bound_min, self._bound_max), - impl=self.tag, - expl=self._expl, - default=self.default, - optional=self.optional, - _decoded=(offset, llen, vlen + (EOC_LEN if lenindef else 0)), - ) + try: + obj = self.__class__( + value=_value, + schema=spec, + bounds=(self._bound_min, self._bound_max), + impl=self.tag, + expl=self._expl, + default=self.default, + optional=self.optional, + _decoded=(offset, llen, vlen + (EOC_LEN if lenindef else 0)), + ) + except BoundsError as err: + raise DecodeError( + msg=str(err), + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) if lenindef: if v[:EOC_LEN].tobytes() != EOC: raise DecodeError( @@ -5045,6 +5273,7 @@ class SequenceOf(Obj): ) obj.lenindef = True tail = v[EOC_LEN:] + obj.ber_encoded = ber_encoded return obj, tail def __repr__(self): @@ -5072,6 +5301,8 @@ 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),)) @@ -5094,6 +5325,16 @@ 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 @@ -5130,10 +5371,21 @@ def generic_decoder(): # pragma: no cover __slots__ = () schema = choice - def pprint_any(obj, oids=None, with_colours=False): + def pprint_any( + obj, + oids=None, + with_colours=False, + with_decode_path=False, + decode_path_only=(), + ): def _pprint_pps(pps): for pp in pps: if hasattr(pp, "_fields"): + if ( + decode_path_only != () and + pp.decode_path[:len(decode_path_only)] != decode_path_only + ): + continue if pp.asn1_type_name == Choice.asn1_type_name: continue pp_kwargs = pp._asdict() @@ -5145,8 +5397,13 @@ def generic_decoder(): # pragma: no cover with_offsets=True, with_blob=False, with_colours=with_colours, + with_decode_path=with_decode_path, + decode_path_len_decrease=len(decode_path_only), ) - for row in pp_console_blob(pp): + for row in pp_console_blob( + pp, + decode_path_len_decrease=len(decode_path_only), + ): yield row else: for row in _pprint_pps(pp): @@ -5178,9 +5435,23 @@ 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", + 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"), @@ -5197,7 +5468,10 @@ 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) @@ -5205,6 +5479,11 @@ def main(): # pragma: no cover obj, oids=oids, with_colours=True if environ.get("NO_COLOR") is None else False, + with_decode_path=args.print_decode_path, + decode_path_only=( + () if args.decode_path_only is None else + tuple(args.decode_path_only.split(":")) + ), )) if tail != b"": print("\nTrailing data: %s" % hexenc(tail))