X-Git-Url: http://www.git.cypherpunks.ru/?a=blobdiff_plain;f=pyderasn.py;h=7ce0565e5da91bd6d57c454752f0a329bc17e3d9;hb=a8f8eb8f0bf0395b8fc94a1cad063def57729a90;hp=0d2d3f5a2982d27fbe49239139c7a01919469104;hpb=a41b107369d403a134c34813ed88ea6808590640;p=pyderasn.git diff --git a/pyderasn.py b/pyderasn.py index 0d2d3f5..7ce0565 100755 --- a/pyderasn.py +++ b/pyderasn.py @@ -1,7 +1,9 @@ #!/usr/bin/env python # coding: utf-8 # cython: language_level=3 -# PyDERASN -- Python ASN.1 DER/BER codec with abstract structures +# pylint: disable=line-too-long,superfluous-parens,protected-access,too-many-lines +# pylint: disable=too-many-return-statements,too-many-branches,too-many-statements +# PyDERASN -- Python ASN.1 DER/CER/BER codec with abstract structures # Copyright (C) 2017-2020 Sergey Matveev # # This program is free software: you can redistribute it and/or modify @@ -75,9 +77,17 @@ tags simultaneously. There are :py:func:`pyderasn.tag_ctxp` and :py:func:`pyderasn.tag_ctxc` functions, allowing you to easily create ``CONTEXT`` ``PRIMITIVE``/``CONSTRUCTED`` tags, by specifying only the required tag -number. Pay attention that explicit tags always have *constructed* tag -(``tag_ctxc``), but implicit tags for primitive types are primitive -(``tag_ctxp``). +number. + +.. note:: + + EXPLICIT tags always have **constructed** tag. PyDERASN does not + explicitly check correctness of schema input here. + +.. note:: + + Implicit tags have **primitive** (``tag_ctxp``) encoding for + primitive values. :: @@ -224,6 +234,7 @@ Currently available context options: * :ref:`allow_unordered_set ` * :ref:`bered ` * :ref:`defines_by_path ` +* :ref:`evgen_mode_upto ` .. _pprinting: @@ -344,7 +355,6 @@ Let's parse that output, human:: (and its derivatives), ``SET``, ``SET OF``, ``UTCTime``, ``GeneralizedTime`` could be BERed. - .. _definedby: DEFINED BY @@ -420,8 +430,8 @@ defines_by_path context option ______________________________ Sometimes you either can not or do not want to explicitly set *defines* -in the scheme. You can dynamically apply those definitions when calling -``.decode()`` method. +in the schema. You can dynamically apply those definitions when calling +:py:meth:`pyderasn.Obj.decode` method. Specify ``defines_by_path`` key in the :ref:`decode context `. Its value must be sequence of following tuples:: @@ -485,9 +495,9 @@ useful for SEQUENCE/SET OF-s. 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 +By default PyDERASN accepts only DER encoded data. By default it 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 ``ber_encoded`` @@ -526,6 +536,377 @@ lengths will be invalid in that case. This option should be used only for skipping some decode errors, just to see the decoded structure somehow. +.. _streaming: + +Streaming and dealing with huge structures +------------------------------------------ + +.. _evgen_mode: + +evgen mode +__________ + +ASN.1 structures can be huge, they can hold millions of objects inside +(for example Certificate Revocation Lists (CRL), holding revocation +state for every previously issued X.509 certificate). CACert.org's 8 MiB +CRL file takes more than half a gigabyte of memory to hold the decoded +structure. + +If you just simply want to check the signature over the ``tbsCertList``, +you can create specialized schema with that field represented as +OctetString for example:: + + class TBSCertListFast(Sequence): + schema = ( + [...] + ("revokedCertificates", OctetString( + impl=SequenceOf.tag_default, + optional=True, + )), + [...] + ) + +This allows you to quickly decode a few fields and check the signature +over the ``tbsCertList`` bytes. + +But how can you get all certificate's serial number from it, after you +trust that CRL after signature validation? You can use so called +``evgen`` (event generation) mode, to catch the events/facts of some +successful object decoding. Let's use command line capabilities:: + + $ python -m pyderasn --schema tests.test_crl:CertificateList --evgen revoke.crl + 10 [1,1, 1] . . version: Version INTEGER v2 (01) OPTIONAL + 15 [1,1, 9] . . . algorithm: OBJECT IDENTIFIER 1.2.840.113549.1.1.13 + 26 [0,0, 2] . . . parameters: [UNIV 5] ANY OPTIONAL + 13 [1,1, 13] . . signature: AlgorithmIdentifier SEQUENCE + 34 [1,1, 3] . . . . . . type: AttributeType OBJECT IDENTIFIER 2.5.4.10 + 39 [0,0, 9] . . . . . . value: [UNIV 19] AttributeValue ANY + 32 [1,1, 14] . . . . . 0: AttributeTypeAndValue SEQUENCE + 30 [1,1, 16] . . . . 0: RelativeDistinguishedName SET OF + [...] + 188 [1,1, 1] . . . . userCertificate: CertificateSerialNumber INTEGER 17 (11) + 191 [1,1, 13] . . . . . utcTime: UTCTime UTCTime 2003-04-01T14:25:08 + 191 [0,0, 15] . . . . revocationDate: Time CHOICE utcTime + 191 [1,1, 13] . . . . . utcTime: UTCTime UTCTime 2003-04-01T14:25:08 + 186 [1,1, 18] . . . 0: RevokedCertificate SEQUENCE + 208 [1,1, 1] . . . . userCertificate: CertificateSerialNumber INTEGER 20 (14) + 211 [1,1, 13] . . . . . utcTime: UTCTime UTCTime 2002-10-01T02:18:01 + 211 [0,0, 15] . . . . revocationDate: Time CHOICE utcTime + 211 [1,1, 13] . . . . . utcTime: UTCTime UTCTime 2002-10-01T02:18:01 + 206 [1,1, 18] . . . 1: RevokedCertificate SEQUENCE + [...] + 9144992 [0,0, 15] . . . . revocationDate: Time CHOICE utcTime + 9144992 [1,1, 13] . . . . . utcTime: UTCTime UTCTime 2020-02-08T07:25:06 + 9144985 [1,1, 20] . . . 415755: RevokedCertificate SEQUENCE + 181 [1,4,9144821] . . revokedCertificates: RevokedCertificates SEQUENCE OF OPTIONAL + 5 [1,4,9144997] . tbsCertList: TBSCertList SEQUENCE + 9145009 [1,1, 9] . . algorithm: OBJECT IDENTIFIER 1.2.840.113549.1.1.13 + 9145020 [0,0, 2] . . parameters: [UNIV 5] ANY OPTIONAL + 9145007 [1,1, 13] . signatureAlgorithm: AlgorithmIdentifier SEQUENCE + 9145022 [1,3, 513] . signatureValue: BIT STRING 4096 bits + 0 [1,4,9145534] CertificateList SEQUENCE + +Here we see how decoder works: it decodes SEQUENCE's tag, length, then +decodes underlying values. It can not tell if SEQUENCE is decoded, so +the event of the upper level SEQUENCE is the last one we see. +``version`` field is just a single INTEGER -- it is decoded and event is +fired immediately. Then we see that ``algorithm`` and ``parameters`` +fields are decoded and only after them the ``signature`` SEQUENCE is +fired as a successfully decoded. There are 4 events for each revoked +certificate entry in that CRL: ``userCertificate`` serial number, +``utcTime`` of ``revocationDate`` CHOICE, ``RevokedCertificate`` itself +as a one of entity in ``revokedCertificates`` SEQUENCE OF. + +We can do that in our ordinary Python code and understand where we are +by looking at deterministically generated decode paths (do not forget +about useful ``--print-decode-path`` CLI option). We must use +:py:meth:`pyderasn.Obj.decode_evgen` method, instead of ordinary +:py:meth:`pyderasn.Obj.decode`. It is generator yielding ``(decode_path, +obj, tail)`` tuples:: + + for decode_path, obj, _ in CertificateList().decode_evgen(crl_raw): + if ( + len(decode_path) == 4 and + decode_path[:2] == ("tbsCertList", "revokedCertificates"), + decode_path[3] == "userCertificate" + ): + print("serial number:", int(obj)) + +Virtually it does not take any memory except at least needed for single +object storage. You can easily use that mode to determine required +object ``.offset`` and ``.*len`` to be able to decode it separately, or +maybe verify signature upon it just by taking bytes by ``.offset`` and +``.tlvlen``. + +.. _evgen_mode_upto_ctx: + +evgen_mode_upto +_______________ + +There is full ability to get any kind of data from the CRL in the +example above. However it is not too convenient to get the whole +``RevokedCertificate`` structure, that is pretty lightweight and one may +do not want to disassemble it. You can use ``evgen_mode_upto`` +:ref:`ctx ` option that semantically equals to +:ref:`defines_by_path ` -- list of decode paths +mapped to any non-None value. If specified decode path is met, then any +subsequent objects won't be decoded in evgen mode. That allows us to +parse the CRL above with fully assembled ``RevokedCertificate``:: + + for decode_path, obj, _ in CertificateList().decode_evgen( + crl_raw, + ctx={"evgen_mode_upto": ( + (("tbsCertList", "revokedCertificates", any), True), + )}, + ): + if ( + len(decode_path) == 3 and + decode_path[:2] == ("tbsCertList", "revokedCertificates"), + ): + print("serial number:", int(obj["userCertificate"])) + +.. note:: + + SEQUENCE/SET values with DEFAULT specified are automatically decoded + without evgen mode. + +.. _mmap: + +mmap-ed file +____________ + +POSIX compliant systems have ``mmap`` syscall, giving ability to work +the memory mapped file. You can deal with the file like it was an +ordinary binary string, allowing you not to load it to the memory first. +Also you can use them as an input for OCTET STRING, taking no Python +memory for their storage. + +There is convenient :py:func:`pyderasn.file_mmaped` function that +creates read-only memoryview on the file contents:: + + with open("huge", "rb") as fd: + raw = file_mmaped(fd) + obj = Something.decode(raw) + +.. warning:: + + mmap-ed files in Python2.7 does not implement buffer protocol, so + memoryview won't work on them. + +.. warning:: + + mmap maps the **whole** file. So it plays no role if you seek-ed it + before. Take the slice of the resulting memoryview with required + offset instead. + +.. note:: + + If you use ZFS as underlying storage, then pay attention that + currently most platforms does not deal good with ZFS ARC and ordinary + page cache used for mmaps. It can take twice the necessary size in + the memory: both in page cache and ZFS ARC. + +CER encoding +____________ + +We can parse any kind of data now, but how can we produce files +streamingly, without storing their encoded representation in memory? +SEQUENCE by default encodes in memory all its values, joins them in huge +binary string, just to know the exact size of SEQUENCE's value for +encoding it in TLV. DER requires you to know all exact sizes of the +objects. + +You can use CER encoding mode, that slightly differs from the DER, but +does not require exact sizes knowledge, allowing streaming encoding +directly to some writer/buffer. Just use +:py:meth:`pyderasn.Obj.encode_cer` method, providing the writer where +encoded data will flow:: + + opener = io.open if PY2 else open + with opener("result", "wb") as fd: + obj.encode_cer(fd.write) + +:: + + buf = io.BytesIO() + obj.encode_cer(buf.write) + +If you do not want to create in-memory buffer every time, then you can +use :py:func:`pyderasn.encode_cer` function:: + + data = encode_cer(obj) + +Remember that CER is **not valid** DER in most cases, so you **have to** +use :ref:`bered ` :ref:`ctx ` option during its +decoding. Also currently there is **no** validation that provided CER is +valid one -- you are sure that it has only valid BER encoding. + +.. warning:: + + SET OF values can not be streamingly encoded, because they are + required to be sorted byte-by-byte. Big SET OF values still will take + much memory. Use neither SET nor SET OF values, as modern ASN.1 + also recommends too. + +Do not forget about using :ref:`mmap-ed ` memoryviews for your +OCTET STRINGs! They will be streamingly copied from underlying file to +the buffer using 1 KB chunks. + +Some structures require that some of the elements have to be forcefully +DER encoded. For example ``SignedData`` CMS requires you to encode +``SignedAttributes`` and X.509 certificates in DER form, allowing you to +encode everything else in BER. You can tell any of the structures to be +forcefully encoded in DER during CER encoding, by specifying +``der_forced=True`` attribute:: + + class Certificate(Sequence): + schema = (...) + der_forced = True + + class SignedAttributes(SetOf): + schema = Attribute() + bounds = (1, 32) + der_forced = True + +.. _agg_octet_string: + +agg_octet_string +________________ + +In most cases, huge quantity of binary data is stored as OCTET STRING. +CER encoding splits it on 1 KB chunks. BER allows splitting on various +levels of chunks inclusion:: + + SOME STRING[CONSTRUCTED] + OCTET STRING[CONSTRUCTED] + OCTET STRING[PRIMITIVE] + DATA CHUNK + OCTET STRING[PRIMITIVE] + DATA CHUNK + OCTET STRING[PRIMITIVE] + DATA CHUNK + OCTET STRING[PRIMITIVE] + DATA CHUNK + OCTET STRING[CONSTRUCTED] + OCTET STRING[PRIMITIVE] + DATA CHUNK + OCTET STRING[PRIMITIVE] + DATA CHUNK + OCTET STRING[CONSTRUCTED] + OCTET STRING[CONSTRUCTED] + OCTET STRING[PRIMITIVE] + DATA CHUNK + +You can not just take the offset and some ``.vlen`` of the STRING and +treat it as the payload. If you decode it without +:ref:`evgen mode `, then it will be automatically aggregated +and ``bytes()`` will give the whole payload contents. + +You are forced to use :ref:`evgen mode ` for decoding for +small memory footprint. There is convenient +:py:func:`pyderasn.agg_octet_string` helper for reconstructing the +payload. Let's assume you have got BER/CER encoded ``ContentInfo`` with +huge ``SignedData`` and ``EncapsulatedContentInfo``. Let's calculate the +SHA512 digest of its ``eContent``:: + + fd = open("data.p7m", "rb") + raw = file_mmaped(fd) + ctx = {"bered": True} + for decode_path, obj, _ in ContentInfo().decode_evgen(raw, ctx=ctx): + if decode_path == ("content",): + content = obj + break + else: + raise ValueError("no content found") + hasher_state = sha512() + def hasher(data): + hasher_state.update(data) + return len(data) + evgens = SignedData().decode_evgen( + raw[content.offset:], + offset=content.offset, + ctx=ctx, + ) + agg_octet_string(evgens, ("encapContentInfo", "eContent"), raw, hasher) + fd.close() + digest = hasher_state.digest() + +Simply replace ``hasher`` with some writeable file's ``fd.write`` to +copy the payload (without BER/CER encoding interleaved overhead) in it. +Virtually it won't take memory more than for keeping small structures +and 1 KB binary chunks. + +.. _seqof-iterators: + +SEQUENCE OF iterators +_____________________ + +You can use iterators as a value in :py:class:`pyderasn.SequenceOf` +classes. The only difference with providing the full list of objects, is +that type and bounds checking is done during encoding process. Also +sequence's value will be emptied after encoding, forcing you to set its +value again. + +This is very useful when you have to create some huge objects, like +CRLs, with thousands and millions of entities inside. You can write the +generator taking necessary data from the database and giving the +``RevokedCertificate`` objects. Only binary representation of that +objects will take memory during DER encoding. + +2-pass DER encoding +------------------- + +There is ability to do 2-pass encoding to DER, writing results directly +to specified writer (buffer, file, whatever). It could be 1.5+ times +slower than ordinary encoding, but it takes little memory for 1st pass +state storing. For example, 1st pass state for CACert.org's CRL with +~416K of certificate entries takes nearly 3.5 MB of memory. +``SignedData`` with several gigabyte ``EncapsulatedContentInfo`` takes +nearly 0.5 KB of memory. + +If you use :ref:`mmap-ed ` memoryviews, :ref:`SEQUENCE OF +iterators ` and write directly to opened file, then +there is very small memory footprint. + +1st pass traverses through all the objects of the structure and returns +the size of DER encoded structure, together with 1st pass state object. +That state contains precalculated lengths for various objects inside the +structure. + +:: + + fulllen, state = obj.encode1st() + +2nd pass takes the writer and 1st pass state. It traverses through all +the objects again, but writes their encoded representation to the writer. + +:: + + opener = io.open if PY2 else open + with opener("result", "wb") as fd: + obj.encode2nd(fd.write, iter(state)) + +.. warning:: + + You **MUST NOT** use 1st pass state if anything is changed in the + objects. It is intended to be used immediately after 1st pass is + done! + +If you use :ref:`SEQUENCE OF iterators `, then you +have to reinitialize the values after the 1st pass. And you **have to** +be sure that the iterator gives exactly the same values as previously. +Yes, you have to run your iterator twice -- because this is two pass +encoding mode. + +If you want to encode to the memory, then you can use convenient +:py:func:`pyderasn.encode2pass` helper. + +.. _browser: + +ASN.1 browser +------------- +.. autofunction:: pyderasn.browse + Base Obj -------- .. autoclass:: pyderasn.Obj @@ -542,12 +923,12 @@ _______ Integer _______ .. autoclass:: pyderasn.Integer - :members: __init__ + :members: __init__, named, tohex BitString _________ .. autoclass:: pyderasn.BitString - :members: __init__ + :members: __init__, bit_len, named OctetString ___________ @@ -579,7 +960,7 @@ _____________ PrintableString _______________ .. autoclass:: pyderasn.PrintableString - :members: __init__ + :members: __init__, allow_asterisk, allow_ampersand UTCTime _______ @@ -589,6 +970,7 @@ _______ GeneralizedTime _______________ .. autoclass:: pyderasn.GeneralizedTime + :members: __init__, todatetime Special types ------------- @@ -596,7 +978,7 @@ Special types Choice ______ .. autoclass:: pyderasn.Choice - :members: __init__ + :members: __init__, choice, value PrimitiveTypes ______________ @@ -634,9 +1016,15 @@ Various ------- .. autofunction:: pyderasn.abs_decode_path +.. autofunction:: pyderasn.agg_octet_string +.. autofunction:: pyderasn.ascii_visualize .. autofunction:: pyderasn.colonize_hex +.. autofunction:: pyderasn.encode2pass +.. autofunction:: pyderasn.encode_cer +.. autofunction:: pyderasn.file_mmaped .. autofunction:: pyderasn.hexenc .. autofunction:: pyderasn.hexdec +.. autofunction:: pyderasn.hexdump .. autofunction:: pyderasn.tag_encode .. autofunction:: pyderasn.tag_decode .. autofunction:: pyderasn.tag_ctxp @@ -653,8 +1041,118 @@ Various .. autoclass:: pyderasn.ObjNotReady .. autoclass:: pyderasn.InvalidValueType .. autoclass:: pyderasn.BoundsError + +.. _cmdline: + +Command-line usage +------------------ + +You can decode DER/BER files using command line abilities:: + + $ python -m pyderasn --schema tests.test_crts:Certificate path/to/file + +If there is no schema for your file, then you can try parsing it without, +but of course IMPLICIT tags will often make it impossible. But result is +good enough for the certificate above:: + + $ python -m pyderasn path/to/file + 0 [1,3,1604] . >: SEQUENCE OF + 4 [1,3,1453] . . >: SEQUENCE OF + 8 [0,0, 5] . . . . >: [0] ANY + . . . . . A0:03:02:01:02 + 13 [1,1, 3] . . . . >: INTEGER 61595 + 18 [1,1, 13] . . . . >: SEQUENCE OF + 20 [1,1, 9] . . . . . . >: OBJECT IDENTIFIER 1.2.840.113549.1.1.5 + 31 [1,1, 0] . . . . . . >: NULL + 33 [1,3, 274] . . . . >: SEQUENCE OF + 37 [1,1, 11] . . . . . . >: SET OF + 39 [1,1, 9] . . . . . . . . >: SEQUENCE OF + 41 [1,1, 3] . . . . . . . . . . >: OBJECT IDENTIFIER 2.5.4.6 + 46 [1,1, 2] . . . . . . . . . . >: PrintableString PrintableString ES + [...] + 1409 [1,1, 50] . . . . . . >: SEQUENCE OF + 1411 [1,1, 8] . . . . . . . . >: OBJECT IDENTIFIER 1.3.6.1.5.5.7.1.1 + 1421 [1,1, 38] . . . . . . . . >: OCTET STRING 38 bytes + . . . . . . . . . 30:24:30:22:06:08:2B:06:01:05:05:07:30:01:86:16 + . . . . . . . . . 68:74:74:70:3A:2F:2F:6F:63:73:70:2E:69:70:73:63 + . . . . . . . . . 61:2E:63:6F:6D:2F + 1461 [1,1, 13] . . >: SEQUENCE OF + 1463 [1,1, 9] . . . . >: OBJECT IDENTIFIER 1.2.840.113549.1.1.5 + 1474 [1,1, 0] . . . . >: NULL + 1476 [1,2, 129] . . >: BIT STRING 1024 bits + . . . 68:EE:79:97:97:DD:3B:EF:16:6A:06:F2:14:9A:6E:CD + . . . 9E:12:F7:AA:83:10:BD:D1:7C:98:FA:C7:AE:D4:0E:2C + [...] + +Human readable OIDs +___________________ + +If you have got dictionaries with ObjectIdentifiers, like example one +from ``tests/test_crts.py``:: + + stroid2name = { + "1.2.840.113549.1.1.1": "id-rsaEncryption", + "1.2.840.113549.1.1.5": "id-sha1WithRSAEncryption", + [...] + "2.5.4.10": "id-at-organizationName", + "2.5.4.11": "id-at-organizationalUnitName", + } + +then you can pass it to pretty printer to see human readable OIDs:: + + $ python -m pyderasn --oids tests.test_crts:stroid2name path/to/file + [...] + 37 [1,1, 11] . . . . . . >: SET OF + 39 [1,1, 9] . . . . . . . . >: SEQUENCE OF + 41 [1,1, 3] . . . . . . . . . . >: OBJECT IDENTIFIER id-at-countryName (2.5.4.6) + 46 [1,1, 2] . . . . . . . . . . >: PrintableString PrintableString ES + 50 [1,1, 18] . . . . . . >: SET OF + 52 [1,1, 16] . . . . . . . . >: SEQUENCE OF + 54 [1,1, 3] . . . . . . . . . . >: OBJECT IDENTIFIER id-at-stateOrProvinceName (2.5.4.8) + 59 [1,1, 9] . . . . . . . . . . >: PrintableString PrintableString Barcelona + 70 [1,1, 18] . . . . . . >: SET OF + 72 [1,1, 16] . . . . . . . . >: SEQUENCE OF + 74 [1,1, 3] . . . . . . . . . . >: OBJECT IDENTIFIER id-at-localityName (2.5.4.7) + 79 [1,1, 9] . . . . . . . . . . >: PrintableString PrintableString Barcelona + [...] + +Decode paths +____________ + +Each decoded element has so-called decode path: sequence of structure +names it is passing during the decode process. Each element has its own +unique path inside the whole ASN.1 tree. You can print it out with +``--print-decode-path`` option:: + + $ python -m pyderasn --schema path.to:Certificate --print-decode-path path/to/file + 0 [1,3,1604] Certificate SEQUENCE [] + 4 [1,3,1453] . tbsCertificate: TBSCertificate SEQUENCE [tbsCertificate] + 10-2 [1,1, 1] . . version: [0] EXPLICIT Version INTEGER v3 OPTIONAL [tbsCertificate:version] + 13 [1,1, 3] . . serialNumber: CertificateSerialNumber INTEGER 61595 [tbsCertificate:serialNumber] + 18 [1,1, 13] . . signature: AlgorithmIdentifier SEQUENCE [tbsCertificate:signature] + 20 [1,1, 9] . . . algorithm: OBJECT IDENTIFIER 1.2.840.113549.1.1.5 [tbsCertificate:signature:algorithm] + 31 [0,0, 2] . . . parameters: [UNIV 5] ANY OPTIONAL [tbsCertificate:signature:parameters] + . . . . 05:00 + 33 [0,0, 278] . . issuer: Name CHOICE rdnSequence [tbsCertificate:issuer] + 33 [1,3, 274] . . . rdnSequence: RDNSequence SEQUENCE OF [tbsCertificate:issuer:rdnSequence] + 37 [1,1, 11] . . . . 0: RelativeDistinguishedName SET OF [tbsCertificate:issuer:rdnSequence:0] + 39 [1,1, 9] . . . . . 0: AttributeTypeAndValue SEQUENCE [tbsCertificate:issuer:rdnSequence:0:0] + 41 [1,1, 3] . . . . . . type: AttributeType OBJECT IDENTIFIER 2.5.4.6 [tbsCertificate:issuer:rdnSequence:0:0:type] + 46 [0,0, 4] . . . . . . value: [UNIV 19] AttributeValue ANY [tbsCertificate:issuer:rdnSequence:0:0:value] + . . . . . . . 13:02:45:53 + 46 [1,1, 2] . . . . . . . DEFINED BY 2.5.4.6: CountryName PrintableString ES [tbsCertificate:issuer:rdnSequence:0:0:value:DEFINED BY 2.5.4.6] + [...] + +Now you can print only the specified tree, for example signature algorithm:: + + $ python -m pyderasn --schema path.to:Certificate --decode-path-only tbsCertificate:signature path/to/file + 18 [1,1, 13] AlgorithmIdentifier SEQUENCE + 20 [1,1, 9] . algorithm: OBJECT IDENTIFIER 1.2.840.113549.1.1.5 + 31 [0,0, 2] . parameters: [UNIV 5] ANY OPTIONAL + . . 05:00 """ +from array import array from codecs import getdecoder from codecs import getencoder from collections import namedtuple @@ -662,10 +1160,15 @@ from collections import OrderedDict from copy import copy from datetime import datetime from datetime import timedelta +from io import BytesIO from math import ceil -from os import environ +from mmap import mmap +from mmap import PROT_READ +from operator import attrgetter from string import ascii_letters from string import digits +from sys import maxsize as sys_maxsize +from sys import version_info from unicodedata import category as unicat from six import add_metaclass @@ -690,19 +1193,24 @@ except ImportError: # pragma: no cover def colored(what, *args, **kwargs): return what -__version__ = "6.2" +__version__ = "7.4" __all__ = ( + "agg_octet_string", "Any", "BitString", "BMPString", "Boolean", "BoundsError", "Choice", + "colonize_hex", "DecodeError", "DecodePathDefBy", + "encode2pass", + "encode_cer", "Enumerated", "ExceedingData", + "file_mmaped", "GeneralizedTime", "GeneralString", "GraphicString", @@ -764,11 +1272,20 @@ EOC = b"\x00\x00" EOC_LEN = len(EOC) LENINDEF = b"\x80" # length indefinite mark LENINDEF_PP_CHAR = "I" if PY2 else "∞" -NAMEDTUPLE_KWARGS = {} if PY2 else {"module": __name__} +NAMEDTUPLE_KWARGS = {} if version_info < (3, 6) else {"module": __name__} SET01 = frozenset("01") DECIMALS = frozenset(digits) DECIMAL_SIGNS = ".," +NEXT_ATTR_NAME = "next" if PY2 else "__next__" + +def file_mmaped(fd): + """Make mmap-ed memoryview for reading from file + + :param fd: file object + :returns: memoryview over read-only mmap-ing of the whole file + """ + return memoryview(mmap(fd.fileno(), 0, prot=PROT_READ)) def pureint(value): if not set(value) <= DECIMALS: @@ -780,6 +1297,19 @@ def fractions2float(fractions_raw): return float("0." + fractions_raw) +def get_def_by_path(defines_by_path, sub_decode_path): + """Get define by decode path + """ + for path, define in defines_by_path: + if len(path) != len(sub_decode_path): + continue + for p1, p2 in zip(path, sub_decode_path): + if (not p1 is any) and (p1 != p2): + break + else: + return define + + ######################################################################## # Errors ######################################################################## @@ -1059,6 +1589,62 @@ def len_decode(data): return l, 1 + octets_num, data[1 + octets_num:] +LEN0 = len_encode(0) +LEN1 = len_encode(1) +LEN1K = len_encode(1000) + + +def len_size(l): + """How many bytes length field will take + """ + if l < 128: + return 1 + if l < 256: # 1 << 8 + return 2 + if l < 65536: # 1 << 16 + return 3 + if l < 16777216: # 1 << 24 + return 4 + if l < 4294967296: # 1 << 32 + return 5 + if l < 1099511627776: # 1 << 40 + return 6 + if l < 281474976710656: # 1 << 48 + return 7 + if l < 72057594037927936: # 1 << 56 + return 8 + raise OverflowError("too big length") + + +def write_full(writer, data): + """Fully write provided data + + :param writer: must comply with ``io.RawIOBase.write`` behaviour + + BytesIO does not guarantee that the whole data will be written at + once. That function write everything provided, raising an error if + ``writer`` returns None. + """ + data = memoryview(data) + written = 0 + while written != len(data): + n = writer(data[written:]) + if n is None: + raise ValueError("can not write to buf") + written += n + + +# If it is 64-bit system, then use compact 64-bit array of unsigned +# longs. Use an ordinary list with universal integers otherwise, that +# is slower. +if sys_maxsize > 2 ** 32: + def state_2pass_new(): + return array("L") +else: + def state_2pass_new(): + return [] + + ######################################################################## # Base class ######################################################################## @@ -1069,6 +1655,22 @@ class AutoAddSlots(type): return type.__new__(cls, name, bases, _dict) +BasicState = namedtuple("BasicState", ( + "version", + "tag", + "tag_order", + "expl", + "default", + "optional", + "offset", + "llen", + "vlen", + "expl_lenindef", + "lenindef", + "ber_encoded", +), **NAMEDTUPLE_KWARGS) + + @add_metaclass(AutoAddSlots) class Obj(object): """Common ASN.1 object class @@ -1078,6 +1680,7 @@ class Obj(object): """ __slots__ = ( "tag", + "_tag_order", "_value", "_expl", "default", @@ -1102,6 +1705,13 @@ class Obj(object): 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") + if self.tag is None: + self._tag_order = None + else: + tag_class, _, tag_num = tag_decode( + self.tag if self._expl is None else self._expl + ) + self._tag_order = (tag_class, tag_num) if default is not None: optional = True self.optional = optional @@ -1141,27 +1751,37 @@ class Obj(object): def __setstate__(self, state): if state.version != __version__: raise ValueError("data is pickled by different PyDERASN version") - self.tag = self.tag_default - self._value = None - self._expl = None - self.default = None - self.optional = False - self.offset = 0 - self.llen = 0 - self.vlen = 0 - self.expl_lenindef = False - self.lenindef = False - self.ber_encoded = False + self.tag = state.tag + self._tag_order = state.tag_order + self._expl = state.expl + self.default = state.default + self.optional = state.optional + self.offset = state.offset + self.llen = state.llen + self.vlen = state.vlen + self.expl_lenindef = state.expl_lenindef + self.lenindef = state.lenindef + self.ber_encoded = state.ber_encoded + + @property + def tag_order(self): + """Tag's (class, number) used for DER/CER sorting + """ + return self._tag_order + + @property + def tag_order_cer(self): + return self.tag_order @property def tlen(self): - """See :ref:`decoding` + """.. seealso:: :ref:`decoding` """ return len(self.tag) @property def tlvlen(self): - """See :ref:`decoding` + """.. seealso:: :ref:`decoding` """ return self.tlen + self.llen + self.vlen @@ -1183,11 +1803,20 @@ class Obj(object): def _encode(self): # pragma: no cover raise NotImplementedError() - def _decode(self, tlv, offset, decode_path, ctx, tag_only): # pragma: no cover + def _encode_cer(self, writer): + write_full(writer, self._encode()) + + def _decode(self, tlv, offset, decode_path, ctx, tag_only, evgen_mode): # pragma: no cover + yield NotImplemented + + def _encode1st(self, state): + raise NotImplementedError() + + def _encode2nd(self, writer, state_iter): raise NotImplementedError() def encode(self): - """Encode the structure + """DER encode the structure :returns: DER representation """ @@ -1196,6 +1825,53 @@ class Obj(object): return raw return b"".join((self._expl, len_encode(len(raw)), raw)) + def encode1st(self, state=None): + """Do the 1st pass of 2-pass encoding + + :rtype: (int, array("L")) + :returns: full length of encoded data and precalculated various + objects lengths + """ + if state is None: + state = state_2pass_new() + if self._expl is None: + return self._encode1st(state) + state.append(0) + idx = len(state) - 1 + vlen, _ = self._encode1st(state) + state[idx] = vlen + fulllen = len(self._expl) + len_size(vlen) + vlen + return fulllen, state + + def encode2nd(self, writer, state_iter): + """Do the 2nd pass of 2-pass encoding + + :param writer: must comply with ``io.RawIOBase.write`` behaviour + :param state_iter: iterator over the 1st pass state (``iter(state)``) + """ + if self._expl is None: + self._encode2nd(writer, state_iter) + else: + write_full(writer, self._expl + len_encode(next(state_iter))) + self._encode2nd(writer, state_iter) + + def encode_cer(self, writer): + """CER encode the structure to specified writer + + :param writer: must comply with ``io.RawIOBase.write`` + behaviour. It takes slice to be written and + returns number of bytes processed. If it returns + None, then exception will be raised + """ + if self._expl is not None: + write_full(writer, self._expl + LENINDEF) + if getattr(self, "der_forced", False): + write_full(writer, self._encode()) + else: + self._encode_cer(writer) + if self._expl is not None: + write_full(writer, EOC) + def hexencode(self): """Do hexadecimal encoded :py:meth:`pyderasn.Obj.encode` """ @@ -1217,32 +1893,78 @@ 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 decode_path: current decode path (tuples of strings, + possibly with DecodePathDefBy) with will be + the root for all underlying objects :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.copy()`` ``ctx`` - before using it? + :param bool tag_only: decode only the tag, without length and + contents (used only in Choice and Set + structures, trying to determine if tag satisfies + the schema) + :param bool _ctx_immutable: do we need to ``copy.copy()`` ``ctx`` + before using it? :returns: (Obj, remaining data) .. seealso:: :ref:`decoding` """ + result = next(self.decode_evgen( + data, + offset, + leavemm, + decode_path, + ctx, + tag_only, + _ctx_immutable, + _evgen_mode=False, + )) + if result is None: + return None + _, obj, tail = result + return obj, tail + + def decode_evgen( + self, + data, + offset=0, + leavemm=False, + decode_path=(), + ctx=None, + tag_only=False, + _ctx_immutable=True, + _evgen_mode=True, + ): + """Decode with evgen mode on + + That method is identical to :py:meth:`pyderasn.Obj.decode`, but + it returns the generator producing ``(decode_path, obj, tail)`` + values. + .. seealso:: :ref:`evgen mode `. + """ if ctx is None: ctx = {} elif _ctx_immutable: ctx = copy(ctx) tlv = memoryview(data) + if ( + _evgen_mode and + get_def_by_path(ctx.get("evgen_mode_upto", ()), decode_path) is not None + ): + _evgen_mode = False if self._expl is None: - result = self._decode( - tlv, - offset, - decode_path=decode_path, - ctx=ctx, - tag_only=tag_only, - ) - if tag_only: - return None - obj, tail = result + for result in self._decode( + tlv, + offset=offset, + decode_path=decode_path, + ctx=ctx, + tag_only=tag_only, + evgen_mode=_evgen_mode, + ): + if tag_only: + yield None + return + _decode_path, obj, tail = result + if not _decode_path is decode_path: + yield result else: try: t, tlen, lv = tag_strip(tlv) @@ -1271,16 +1993,20 @@ class Obj(object): ) llen, v = 1, lv[1:] offset += tlen + llen - result = self._decode( - v, - offset=offset, - decode_path=decode_path, - ctx=ctx, - tag_only=tag_only, - ) - if tag_only: # pragma: no cover - return None - obj, tail = result + for result in self._decode( + v, + offset=offset, + decode_path=decode_path, + ctx=ctx, + tag_only=tag_only, + evgen_mode=_evgen_mode, + ): + if tag_only: # pragma: no cover + yield None + return + _decode_path, obj, tail = result + if not _decode_path is decode_path: + yield result eoc_expected, tail = tail[:EOC_LEN], tail[EOC_LEN:] if eoc_expected.tobytes() != EOC: raise DecodeError( @@ -1306,16 +2032,20 @@ class Obj(object): decode_path=decode_path, offset=offset, ) - result = self._decode( - v, - offset=offset + tlen + llen, - decode_path=decode_path, - ctx=ctx, - tag_only=tag_only, - ) - if tag_only: # pragma: no cover - return None - obj, tail = result + for result in self._decode( + v, + offset=offset + tlen + llen, + decode_path=decode_path, + ctx=ctx, + tag_only=tag_only, + evgen_mode=_evgen_mode, + ): + if tag_only: # pragma: no cover + yield None + return + _decode_path, obj, tail = result + if not _decode_path is decode_path: + yield result if obj.tlvlen < l and not ctx.get("allow_expl_oob", False): raise DecodeError( "explicit tag out-of-bound, longer than data", @@ -1323,7 +2053,7 @@ class Obj(object): decode_path=decode_path, offset=offset, ) - return obj, (tail if leavemm else tail.tobytes()) + yield decode_path, obj, (tail if leavemm else tail.tobytes()) def decod(self, data, offset=0, decode_path=(), ctx=None): """Decode the data, check that tail is empty @@ -1357,25 +2087,25 @@ class Obj(object): @property def expled(self): - """See :ref:`decoding` + """.. seealso:: :ref:`decoding` """ return self._expl is not None @property def expl_tag(self): - """See :ref:`decoding` + """.. seealso:: :ref:`decoding` """ return self._expl @property def expl_tlen(self): - """See :ref:`decoding` + """.. seealso:: :ref:`decoding` """ return len(self._expl) @property def expl_llen(self): - """See :ref:`decoding` + """.. seealso:: :ref:`decoding` """ if self.expl_lenindef: return 1 @@ -1383,31 +2113,31 @@ class Obj(object): @property def expl_offset(self): - """See :ref:`decoding` + """.. seealso:: :ref:`decoding` """ return self.offset - self.expl_tlen - self.expl_llen @property def expl_vlen(self): - """See :ref:`decoding` + """.. seealso:: :ref:`decoding` """ return self.tlvlen @property def expl_tlvlen(self): - """See :ref:`decoding` + """.. seealso:: :ref:`decoding` """ return self.expl_tlen + self.expl_llen + self.expl_vlen @property def fulloffset(self): - """See :ref:`decoding` + """.. seealso:: :ref:`decoding` """ return self.expl_offset if self.expled else self.offset @property def fulllen(self): - """See :ref:`decoding` + """.. seealso:: :ref:`decoding` """ return self.expl_tlvlen if self.expled else self.tlvlen @@ -1444,6 +2174,27 @@ class Obj(object): ) +def encode_cer(obj): + """Encode to CER in memory buffer + + :returns bytes: memory buffer contents + """ + buf = BytesIO() + obj.encode_cer(buf.write) + return buf.getvalue() + + +def encode2pass(obj): + """Encode (2-pass mode) to DER in memory buffer + + :returns bytes: memory buffer contents + """ + buf = BytesIO() + _, state = obj.encode1st() + obj.encode2nd(buf.write, iter(state)) + return buf.getvalue() + + class DecodePathDefBy(object): """DEFINED BY representation inside decode path """ @@ -1557,6 +2308,15 @@ def colonize_hex(hexed): return ":".join(hexed[i:i + 2] for i in six_xrange(0, len(hexed), 2)) +def find_oid_name(asn1_type_name, oid_maps, value): + if len(oid_maps) > 0 and asn1_type_name == ObjectIdentifier.asn1_type_name: + for oid_map in oid_maps: + oid_name = oid_map.get(value) + if oid_name is not None: + return oid_name + return None + + def pp_console_row( pp, oid_maps=(), @@ -1580,9 +2340,7 @@ def pp_console_row( col += _colourize("B", "red", with_colours) if pp.bered else " " cols.append(col) col = "[%d,%d,%4d]%s" % ( - pp.tlen, - pp.llen, - pp.vlen, + pp.tlen, pp.llen, pp.vlen, LENINDEF_PP_CHAR if pp.lenindef else " " ) col = _colourize(col, "green", with_colours, ()) @@ -1594,19 +2352,11 @@ def pp_console_row( if isinstance(ent, DecodePathDefBy): cols.append(_colourize("DEFINED BY", "red", with_colours, ("reverse",))) value = str(ent.defined_by) - oid_name = None - if ( - len(oid_maps) > 0 and - ent.defined_by.asn1_type_name == - ObjectIdentifier.asn1_type_name - ): - for oid_map in oid_maps: - oid_name = oid_map.get(value) - if oid_name is not None: - cols.append(_colourize("%s:" % oid_name, "green", with_colours)) - break + oid_name = find_oid_name(ent.defined_by.asn1_type_name, oid_maps, value) if oid_name is None: cols.append(_colourize("%s:" % value, "white", with_colours, ("reverse",))) + else: + cols.append(_colourize("%s:" % oid_name, "green", with_colours)) else: cols.append(_colourize("%s:" % ent, "yellow", with_colours, ("reverse",))) if pp.expl is not None: @@ -1625,23 +2375,12 @@ def pp_console_row( if pp.value is not None: value = pp.value cols.append(_colourize(value, "white", with_colours, ("reverse",))) - if ( - len(oid_maps) > 0 and - pp.asn1_type_name == ObjectIdentifier.asn1_type_name - ): - for oid_map in oid_maps: - oid_name = oid_map.get(value) - if oid_name is not None: - cols.append(_colourize("(%s)" % oid_name, "green", with_colours)) - break + oid_name = find_oid_name(pp.asn1_type_name, oid_maps, pp.value) + if oid_name is not None: + cols.append(_colourize("(%s)" % oid_name, "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, + "(%s)" % colonize_hex(pp.obj.tohex()), "green", with_colours, )) if with_blob: if pp.blob.__class__ == binary_type: @@ -1682,11 +2421,12 @@ def pprint( with_colours=False, with_decode_path=False, decode_path_only=(), + decode_path=(), ): """Pretty print object :param Obj obj: object you want to pretty print - :param oid_maps: list of ``str(OID) <-> human readable string`` dictionary. + :param oid_maps: list of ``str(OID) <-> human readable string`` dictionaries. Its human readable form is printed when OID is met :param big_blobs: if large binary objects are met (like OctetString values), do we need to print them too, on separate @@ -1734,27 +2474,18 @@ def pprint( else: for row in _pprint_pps(pp): yield row - return "\n".join(_pprint_pps(obj.pps())) + return "\n".join(_pprint_pps(obj.pps(decode_path))) ######################################################################## # ASN.1 primitive types ######################################################################## -BooleanState = namedtuple("BooleanState", ( - "version", - "value", - "tag", - "expl", - "default", - "optional", - "offset", - "llen", - "vlen", - "expl_lenindef", - "lenindef", - "ber_encoded", -), **NAMEDTUPLE_KWARGS) +BooleanState = namedtuple( + "BooleanState", + BasicState._fields + ("value",), + **NAMEDTUPLE_KWARGS +) class Boolean(Obj): @@ -1814,8 +2545,8 @@ class Boolean(Obj): def __getstate__(self): return BooleanState( __version__, - self._value, self.tag, + self._tag_order, self._expl, self.default, self.optional, @@ -1825,21 +2556,12 @@ class Boolean(Obj): self.expl_lenindef, self.lenindef, self.ber_encoded, + self._value, ) def __setstate__(self, state): super(Boolean, self).__setstate__(state) self._value = state.value - self.tag = state.tag - self._expl = state.expl - self.default = state.default - self.optional = state.optional - self.offset = state.offset - self.llen = state.llen - self.vlen = state.vlen - self.expl_lenindef = state.expl_lenindef - self.lenindef = state.lenindef - self.ber_encoded = state.ber_encoded def __nonzero__(self): self._assert_ready() @@ -1878,13 +2600,16 @@ class Boolean(Obj): def _encode(self): self._assert_ready() - return b"".join(( - self.tag, - len_encode(1), - (b"\xFF" if self._value else b"\x00"), - )) + return b"".join((self.tag, LEN1, (b"\xFF" if self._value else b"\x00"))) - def _decode(self, tlv, offset, decode_path, ctx, tag_only): + def _encode1st(self, state): + return len(self.tag) + 2, state + + def _encode2nd(self, writer, state_iter): + self._assert_ready() + write_full(writer, self._encode()) + + def _decode(self, tlv, offset, decode_path, ctx, tag_only, evgen_mode): try: t, _, lv = tag_strip(tlv) except DecodeError as err: @@ -1901,7 +2626,8 @@ class Boolean(Obj): offset=offset, ) if tag_only: - return None + yield None + return try: l, _, v = len_decode(lv) except DecodeError as err: @@ -1950,7 +2676,7 @@ class Boolean(Obj): _decoded=(offset, 1, 1), ) obj.ber_encoded = ber_encoded - return obj, v[1:] + yield decode_path, obj, v[1:] def __repr__(self): return pp_console_row(next(self.pps())) @@ -1982,23 +2708,11 @@ class Boolean(Obj): yield pp -IntegerState = namedtuple("IntegerState", ( - "version", - "specs", - "value", - "bound_min", - "bound_max", - "tag", - "expl", - "default", - "optional", - "offset", - "llen", - "vlen", - "expl_lenindef", - "lenindef", - "ber_encoded", -), **NAMEDTUPLE_KWARGS) +IntegerState = namedtuple( + "IntegerState", + BasicState._fields + ("specs", "value", "bound_min", "bound_max"), + **NAMEDTUPLE_KWARGS +) class Integer(Obj): @@ -2105,11 +2819,8 @@ class Integer(Obj): def __getstate__(self): return IntegerState( __version__, - self.specs, - self._value, - self._bound_min, - self._bound_max, self.tag, + self._tag_order, self._expl, self.default, self.optional, @@ -2119,6 +2830,10 @@ class Integer(Obj): self.expl_lenindef, self.lenindef, self.ber_encoded, + self.specs, + self._value, + self._bound_min, + self._bound_max, ) def __setstate__(self, state): @@ -2127,28 +2842,28 @@ class Integer(Obj): self._value = state.value self._bound_min = state.bound_min self._bound_max = state.bound_max - self.tag = state.tag - self._expl = state.expl - self.default = state.default - self.optional = state.optional - self.offset = state.offset - self.llen = state.llen - self.vlen = state.vlen - self.expl_lenindef = state.expl_lenindef - self.lenindef = state.lenindef - self.ber_encoded = state.ber_encoded def __int__(self): self._assert_ready() return int(self._value) + def tohex(self): + """Hexadecimal representation + + Use :py:func:`pyderasn.colonize_hex` for colonizing it. + """ + hex_repr = hex(int(self))[2:].upper() + if len(hex_repr) % 2 != 0: + hex_repr = "0" + hex_repr + return hex_repr + def __hash__(self): self._assert_ready() - return hash( - self.tag + - bytes(self._expl or b"") + + return hash(b"".join(( + self.tag, + bytes(self._expl or b""), str(self._value).encode("ascii"), - ) + ))) def __eq__(self, their): if isinstance(their, integer_types): @@ -2166,6 +2881,8 @@ class Integer(Obj): @property def named(self): + """Return named representation (if exists) of the value + """ for name, value in iteritems(self.specs): if value == self._value: return name @@ -2193,7 +2910,7 @@ class Integer(Obj): _specs=self.specs, ) - def _encode(self): + def _encode_payload(self): self._assert_ready() value = self._value if PY2: @@ -2230,9 +2947,20 @@ class Integer(Obj): bytes_len += 1 else: break + return octets + + def _encode(self): + octets = self._encode_payload() return b"".join((self.tag, len_encode(len(octets)), octets)) - def _decode(self, tlv, offset, decode_path, ctx, tag_only): + def _encode1st(self, state): + l = len(self._encode_payload()) + return len(self.tag) + len_size(l) + l, state + + def _encode2nd(self, writer, state_iter): + write_full(writer, self._encode()) + + def _decode(self, tlv, offset, decode_path, ctx, tag_only, evgen_mode): try: t, _, lv = tag_strip(tlv) except DecodeError as err: @@ -2249,7 +2977,8 @@ class Integer(Obj): offset=offset, ) if tag_only: - return None + yield None + return try: l, llen, v = len_decode(lv) except DecodeError as err: @@ -2320,7 +3049,7 @@ class Integer(Obj): decode_path=decode_path, offset=offset, ) - return obj, tail + yield decode_path, obj, tail def __repr__(self): return pp_console_row(next(self.pps())) @@ -2351,23 +3080,11 @@ class Integer(Obj): yield pp -BitStringState = namedtuple("BitStringState", ( - "version", - "specs", - "value", - "tag", - "expl", - "default", - "optional", - "offset", - "llen", - "vlen", - "expl_lenindef", - "lenindef", - "ber_encoded", - "tag_constructed", - "defined", -), **NAMEDTUPLE_KWARGS) +BitStringState = namedtuple( + "BitStringState", + BasicState._fields + ("specs", "value", "tag_constructed", "defined"), + **NAMEDTUPLE_KWARGS +) class BitString(Obj): @@ -2525,9 +3242,8 @@ class BitString(Obj): def __getstate__(self): return BitStringState( __version__, - self.specs, - self._value, self.tag, + self._tag_order, self._expl, self.default, self.optional, @@ -2537,6 +3253,8 @@ class BitString(Obj): self.expl_lenindef, self.lenindef, self.ber_encoded, + self.specs, + self._value, self.tag_constructed, self.defined, ) @@ -2545,16 +3263,6 @@ class BitString(Obj): super(BitString, self).__setstate__(state) self.specs = state.specs self._value = state.value - self.tag = state.tag - self._expl = state.expl - self.default = state.default - self.optional = state.optional - self.offset = state.offset - self.llen = state.llen - self.vlen = state.vlen - self.expl_lenindef = state.expl_lenindef - self.lenindef = state.lenindef - self.ber_encoded = state.ber_encoded self.tag_constructed = state.tag_constructed self.defined = state.defined @@ -2565,6 +3273,8 @@ class BitString(Obj): @property def bit_len(self): + """Returns number of bits in the string + """ self._assert_ready() return self._value[0] @@ -2585,6 +3295,10 @@ class BitString(Obj): @property def named(self): + """Named representation (if exists) of the bits + + :returns: [str(name), ...] + """ return [name for name, bit in iteritems(self.specs) if self[bit]] def __call__( @@ -2630,65 +3344,46 @@ class BitString(Obj): octets, )) - def _decode_chunk(self, lv, offset, decode_path): - try: - l, llen, v = len_decode(lv) - except DecodeError as err: - raise err.__class__( - msg=err.msg, - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) - if l > len(v): - raise NotEnoughData( - "encoded length is longer than data", - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) - if l == 0: - raise NotEnoughData( - "zero length", - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) - pad_size = byte2int(v) - if l == 1 and pad_size != 0: - raise DecodeError( - "invalid empty value", - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) - if pad_size > 7: - raise DecodeError( - "too big pad", - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) - if byte2int(v[l - 1:l]) & ((1 << pad_size) - 1) != 0: - raise DecodeError( - "invalid pad", - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) - v, tail = v[:l], v[l:] - obj = self.__class__( - value=((len(v) - 1) * 8 - pad_size, v[1:].tobytes()), - impl=self.tag, - expl=self._expl, - default=self.default, - optional=self.optional, - _specs=self.specs, - _decoded=(offset, llen, l), - ) - return obj, tail + def _encode1st(self, state): + self._assert_ready() + _, octets = self._value + l = len(octets) + 1 + return len(self.tag) + len_size(l) + l, state + + def _encode2nd(self, writer, state_iter): + bit_len, octets = self._value + write_full(writer, b"".join(( + self.tag, + len_encode(len(octets) + 1), + int2byte((8 - bit_len % 8) % 8), + ))) + write_full(writer, octets) - def _decode(self, tlv, offset, decode_path, ctx, tag_only): + def _encode_cer(self, writer): + bit_len, octets = self._value + if len(octets) + 1 <= 1000: + write_full(writer, self._encode()) + return + write_full(writer, self.tag_constructed) + write_full(writer, LENINDEF) + for offset in six_xrange(0, (len(octets) // 999) * 999, 999): + write_full(writer, b"".join(( + BitString.tag_default, + LEN1K, + int2byte(0), + octets[offset:offset + 999], + ))) + tail = octets[offset+999:] + if len(tail) > 0: + tail = int2byte((8 - bit_len % 8) % 8) + tail + write_full(writer, b"".join(( + BitString.tag_default, + len_encode(len(tail)), + tail, + ))) + write_full(writer, EOC) + + def _decode(self, tlv, offset, decode_path, ctx, tag_only, evgen_mode): try: t, tlen, lv = tag_strip(tlv) except DecodeError as err: @@ -2700,24 +3395,10 @@ class BitString(Obj): ) if t == self.tag: if tag_only: # pragma: no cover - return None - return self._decode_chunk(lv, offset, decode_path) - if t == self.tag_constructed: - if not ctx.get("bered", False): - raise DecodeError( - "unallowed BER constructed encoding", - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) - if tag_only: # pragma: no cover - return None - lenindef = False + yield None + return try: l, llen, v = len_decode(lv) - except LenIndefForm: - llen, l, v = 1, 0, lv[1:] - lenindef = True except DecodeError as err: raise err.__class__( msg=err.msg, @@ -2732,122 +3413,213 @@ class BitString(Obj): decode_path=decode_path, offset=offset, ) - if not lenindef and l == 0: + if 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 lenindef: - if v[:EOC_LEN].tobytes() == EOC: - break - else: - if vlen == l: - break - if vlen > l: - raise DecodeError( - "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)),) - try: - chunk, v_tail = BitString().decode( - v, - offset=sub_offset, - decode_path=sub_decode_path, - leavemm=True, - ctx=ctx, - _ctx_immutable=False, - ) - except TagMismatch: - raise DecodeError( - "expected BitString encoded chunk", - klass=self.__class__, - decode_path=sub_decode_path, - offset=sub_offset, - ) - chunks.append(chunk) - sub_offset += chunk.tlvlen - vlen += chunk.tlvlen - v = v_tail - if len(chunks) == 0: + pad_size = byte2int(v) + if l == 1 and pad_size != 0: raise DecodeError( - "no chunks", + "invalid empty value", klass=self.__class__, decode_path=decode_path, offset=offset, ) - values = [] - bit_len = 0 - for chunk_i, chunk in enumerate(chunks[:-1]): - if chunk.bit_len % 8 != 0: - raise DecodeError( - "BitString chunk is not multiple of 8 bits", - klass=self.__class__, - decode_path=decode_path + (str(chunk_i),), - offset=chunk.offset, - ) - values.append(bytes(chunk)) - bit_len += chunk.bit_len - chunk_last = chunks[-1] - values.append(bytes(chunk_last)) - bit_len += chunk_last.bit_len + if pad_size > 7: + raise DecodeError( + "too big pad", + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + if byte2int(v[l - 1:l]) & ((1 << pad_size) - 1) != 0: + raise DecodeError( + "invalid pad", + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + v, tail = v[:l], v[l:] + bit_len = (len(v) - 1) * 8 - pad_size obj = self.__class__( - value=(bit_len, b"".join(values)), + value=None if evgen_mode else (bit_len, v[1:].tobytes()), impl=self.tag, expl=self._expl, default=self.default, optional=self.optional, _specs=self.specs, - _decoded=(offset, llen, vlen + (EOC_LEN if lenindef else 0)), + _decoded=(offset, llen, l), ) - obj.lenindef = lenindef - obj.ber_encoded = True - return obj, (v[EOC_LEN:] if lenindef else v) - raise TagMismatch( - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) - - def __repr__(self): - return pp_console_row(next(self.pps())) - - def pps(self, decode_path=()): - value = None - blob = None - if self.ready: - bit_len, blob = self._value - value = "%d bits" % bit_len - 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, - value=value, - blob=blob, - optional=self.optional, - default=self == self.default, - impl=None if self.tag == self.tag_default else tag_decode(self.tag), - expl=None if self._expl is None else tag_decode(self._expl), - offset=self.offset, - 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, + if evgen_mode: + obj._value = (bit_len, None) + yield decode_path, obj, tail + return + if t != self.tag_constructed: + raise TagMismatch( + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + if not ctx.get("bered", False): + raise DecodeError( + "unallowed BER constructed encoding", + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + if tag_only: # pragma: no cover + yield None + return + lenindef = False + try: + l, llen, v = len_decode(lv) + except LenIndefForm: + llen, l, v = 1, 0, lv[1:] + lenindef = True + except DecodeError as err: + raise err.__class__( + msg=err.msg, + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + if l > len(v): + raise NotEnoughData( + "encoded length is longer than data", + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + if not lenindef 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 lenindef: + if v[:EOC_LEN].tobytes() == EOC: + break + else: + if vlen == l: + break + if vlen > l: + raise DecodeError( + "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)),) + try: + if evgen_mode: + for _decode_path, chunk, v_tail in BitString().decode_evgen( + v, + offset=sub_offset, + decode_path=sub_decode_path, + leavemm=True, + ctx=ctx, + _ctx_immutable=False, + ): + yield _decode_path, chunk, v_tail + else: + _, chunk, v_tail = next(BitString().decode_evgen( + v, + offset=sub_offset, + decode_path=sub_decode_path, + leavemm=True, + ctx=ctx, + _ctx_immutable=False, + _evgen_mode=False, + )) + except TagMismatch: + raise DecodeError( + "expected BitString encoded chunk", + klass=self.__class__, + decode_path=sub_decode_path, + offset=sub_offset, + ) + chunks.append(chunk) + sub_offset += chunk.tlvlen + vlen += chunk.tlvlen + v = v_tail + if len(chunks) == 0: + raise DecodeError( + "no chunks", + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + values = [] + bit_len = 0 + for chunk_i, chunk in enumerate(chunks[:-1]): + if chunk.bit_len % 8 != 0: + raise DecodeError( + "BitString chunk is not multiple of 8 bits", + klass=self.__class__, + decode_path=decode_path + (str(chunk_i),), + offset=chunk.offset, + ) + if not evgen_mode: + values.append(bytes(chunk)) + bit_len += chunk.bit_len + chunk_last = chunks[-1] + if not evgen_mode: + values.append(bytes(chunk_last)) + bit_len += chunk_last.bit_len + obj = self.__class__( + value=None if evgen_mode else (bit_len, b"".join(values)), + impl=self.tag, + expl=self._expl, + default=self.default, + optional=self.optional, + _specs=self.specs, + _decoded=(offset, llen, vlen + (EOC_LEN if lenindef else 0)), + ) + if evgen_mode: + obj._value = (bit_len, None) + obj.lenindef = lenindef + obj.ber_encoded = True + yield decode_path, obj, (v[EOC_LEN:] if lenindef else v) + + def __repr__(self): + return pp_console_row(next(self.pps())) + + def pps(self, decode_path=()): + value = None + blob = None + if self.ready: + bit_len, blob = self._value + value = "%d bits" % bit_len + if len(self.specs) > 0 and blob is not None: + blob = tuple(self.named) + yield _pp( + obj=self, + asn1_type_name=self.asn1_type_name, + obj_name=self.__class__.__name__, + decode_path=decode_path, + value=value, + blob=blob, + optional=self.optional, + default=self == self.default, + impl=None if self.tag == self.tag_default else tag_decode(self.tag), + expl=None if self._expl is None else tag_decode(self._expl), + offset=self.offset, + 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, lenindef=self.lenindef, ber_encoded=self.ber_encoded, bered=self.bered, @@ -2861,24 +3633,17 @@ class BitString(Obj): yield pp -OctetStringState = namedtuple("OctetStringState", ( - "version", - "value", - "bound_min", - "bound_max", - "tag", - "expl", - "default", - "optional", - "offset", - "llen", - "vlen", - "expl_lenindef", - "lenindef", - "ber_encoded", - "tag_constructed", - "defined", -), **NAMEDTUPLE_KWARGS) +OctetStringState = namedtuple( + "OctetStringState", + BasicState._fields + ( + "value", + "bound_min", + "bound_max", + "tag_constructed", + "defined", + ), + **NAMEDTUPLE_KWARGS +) class OctetString(Obj): @@ -2897,17 +3662,15 @@ class OctetString(Obj): >>> 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. + Memoryviews can be used as a values. If memoryview is made on + mmap-ed file, then it does not take storage inside OctetString + itself. In CER encoding mode it will be streamed to the specified + writer, copying 1 KB chunks. """ __slots__ = ("tag_constructed", "_bound_min", "_bound_max", "defined") tag_default = tag_encode(4) asn1_type_name = "OCTET STRING" + evgen_mode_skip_value = True def __init__( self, @@ -2957,12 +3720,12 @@ class OctetString(Obj): ) def _value_sanitize(self, value): - if value.__class__ == binary_type: + if value.__class__ == binary_type or value.__class__ == memoryview: pass elif issubclass(value.__class__, OctetString): value = value._value else: - raise InvalidValueType((self.__class__, bytes)) + raise InvalidValueType((self.__class__, bytes, memoryview)) if not self._bound_min <= len(value) <= self._bound_max: raise BoundsError(self._bound_min, len(value), self._bound_max) return value @@ -2974,10 +3737,8 @@ class OctetString(Obj): def __getstate__(self): return OctetStringState( __version__, - self._value, - self._bound_min, - self._bound_max, self.tag, + self._tag_order, self._expl, self.default, self.optional, @@ -2987,6 +3748,9 @@ class OctetString(Obj): self.expl_lenindef, self.lenindef, self.ber_encoded, + self._value, + self._bound_min, + self._bound_max, self.tag_constructed, self.defined, ) @@ -2996,22 +3760,12 @@ class OctetString(Obj): self._value = state.value self._bound_min = state.bound_min self._bound_max = state.bound_max - self.tag = state.tag - self._expl = state.expl - self.default = state.default - self.optional = state.optional - self.offset = state.offset - self.llen = state.llen - self.vlen = state.vlen - self.expl_lenindef = state.expl_lenindef - self.lenindef = state.lenindef - self.ber_encoded = state.ber_encoded self.tag_constructed = state.tag_constructed self.defined = state.defined def __bytes__(self): self._assert_ready() - return self._value + return bytes(self._value) def __eq__(self, their): if their.__class__ == binary_type: @@ -3056,52 +3810,39 @@ class OctetString(Obj): self._value, )) - def _decode_chunk(self, lv, offset, decode_path, ctx): - try: - l, llen, v = len_decode(lv) - except DecodeError as err: - raise err.__class__( - msg=err.msg, - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) - if l > len(v): - raise NotEnoughData( - "encoded length is longer than data", - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) - v, tail = v[:l], v[l:] - try: - obj = self.__class__( - value=v.tobytes(), - bounds=(self._bound_min, self._bound_max), - impl=self.tag, - expl=self._expl, - default=self.default, - optional=self.optional, - _decoded=(offset, llen, l), - ctx=ctx, - ) - except DecodeError as err: - raise DecodeError( - msg=err.msg, - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) - except BoundsError as err: - raise DecodeError( - msg=str(err), - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) - return obj, tail + def _encode1st(self, state): + self._assert_ready() + l = len(self._value) + return len(self.tag) + len_size(l) + l, state - def _decode(self, tlv, offset, decode_path, ctx, tag_only): + def _encode2nd(self, writer, state_iter): + value = self._value + write_full(writer, self.tag + len_encode(len(value))) + write_full(writer, value) + + def _encode_cer(self, writer): + octets = self._value + if len(octets) <= 1000: + write_full(writer, self._encode()) + return + write_full(writer, self.tag_constructed) + write_full(writer, LENINDEF) + for offset in six_xrange(0, (len(octets) // 1000) * 1000, 1000): + write_full(writer, b"".join(( + OctetString.tag_default, + LEN1K, + octets[offset:offset + 1000], + ))) + tail = octets[offset+1000:] + if len(tail) > 0: + write_full(writer, b"".join(( + OctetString.tag_default, + len_encode(len(tail)), + tail, + ))) + write_full(writer, EOC) + + def _decode(self, tlv, offset, decode_path, ctx, tag_only, evgen_mode): try: t, tlen, lv = tag_strip(tlv) except DecodeError as err: @@ -3113,24 +3854,10 @@ class OctetString(Obj): ) if t == self.tag: if tag_only: - return None - return self._decode_chunk(lv, offset, decode_path, ctx) - if t == self.tag_constructed: - if not ctx.get("bered", False): - raise DecodeError( - "unallowed BER constructed encoding", - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) - if tag_only: - return None - lenindef = False + yield None + return try: l, llen, v = len_decode(lv) - except LenIndefForm: - llen, l, v = 1, 0, lv[1:] - lenindef = True except DecodeError as err: raise err.__class__( msg=err.msg, @@ -3145,53 +3872,26 @@ class OctetString(Obj): decode_path=decode_path, offset=offset, ) - chunks = [] - sub_offset = offset + tlen + llen - vlen = 0 - while True: - if lenindef: - if v[:EOC_LEN].tobytes() == EOC: - break - else: - if vlen == l: - break - if vlen > l: - raise DecodeError( - "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)),) - try: - chunk, v_tail = OctetString().decode( - v, - offset=sub_offset, - decode_path=sub_decode_path, - leavemm=True, - ctx=ctx, - _ctx_immutable=False, - ) - except TagMismatch: - raise DecodeError( - "expected OctetString encoded chunk", - klass=self.__class__, - decode_path=sub_decode_path, - offset=sub_offset, - ) - chunks.append(chunk) - sub_offset += chunk.tlvlen - vlen += chunk.tlvlen - v = v_tail + v, tail = v[:l], v[l:] + if evgen_mode and not self._bound_min <= len(v) <= self._bound_max: + raise DecodeError( + msg=str(BoundsError(self._bound_min, len(v), self._bound_max)), + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) try: obj = self.__class__( - value=b"".join(bytes(chunk) for chunk in chunks), + value=( + None if (evgen_mode and self.evgen_mode_skip_value) + else v.tobytes() + ), 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)), + _decoded=(offset, llen, l), ctx=ctx, ) except DecodeError as err: @@ -3208,14 +3908,138 @@ class OctetString(Obj): decode_path=decode_path, offset=offset, ) - obj.lenindef = lenindef - obj.ber_encoded = True - return obj, (v[EOC_LEN:] if lenindef else v) - raise TagMismatch( - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) + yield decode_path, obj, tail + return + if t != self.tag_constructed: + raise TagMismatch( + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + if not ctx.get("bered", False): + raise DecodeError( + "unallowed BER constructed encoding", + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + if tag_only: + yield None + return + lenindef = False + try: + l, llen, v = len_decode(lv) + except LenIndefForm: + llen, l, v = 1, 0, lv[1:] + lenindef = True + except DecodeError as err: + raise err.__class__( + msg=err.msg, + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + if l > len(v): + raise NotEnoughData( + "encoded length is longer than data", + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + chunks = [] + chunks_count = 0 + sub_offset = offset + tlen + llen + vlen = 0 + payload_len = 0 + while True: + if lenindef: + if v[:EOC_LEN].tobytes() == EOC: + break + else: + if vlen == l: + break + if vlen > l: + raise DecodeError( + "chunk out of bounds", + klass=self.__class__, + decode_path=decode_path + (str(len(chunks) - 1),), + offset=chunks[-1].offset, + ) + try: + if evgen_mode: + sub_decode_path = decode_path + (str(chunks_count),) + for _decode_path, chunk, v_tail in OctetString().decode_evgen( + v, + offset=sub_offset, + decode_path=sub_decode_path, + leavemm=True, + ctx=ctx, + _ctx_immutable=False, + ): + yield _decode_path, chunk, v_tail + if not chunk.ber_encoded: + payload_len += chunk.vlen + chunks_count += 1 + else: + sub_decode_path = decode_path + (str(len(chunks)),) + _, chunk, v_tail = next(OctetString().decode_evgen( + v, + offset=sub_offset, + decode_path=sub_decode_path, + leavemm=True, + ctx=ctx, + _ctx_immutable=False, + _evgen_mode=False, + )) + chunks.append(chunk) + except TagMismatch: + raise DecodeError( + "expected OctetString encoded chunk", + klass=self.__class__, + decode_path=sub_decode_path, + offset=sub_offset, + ) + sub_offset += chunk.tlvlen + vlen += chunk.tlvlen + v = v_tail + if evgen_mode and not self._bound_min <= payload_len <= self._bound_max: + raise DecodeError( + msg=str(BoundsError(self._bound_min, payload_len, self._bound_max)), + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + try: + obj = self.__class__( + value=( + None if evgen_mode else + b"".join(bytes(chunk) for chunk in chunks) + ), + 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)), + ctx=ctx, + ) + except DecodeError as err: + raise DecodeError( + msg=err.msg, + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + except BoundsError as err: + raise DecodeError( + msg=str(err), + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + obj.lenindef = lenindef + obj.ber_encoded = True + yield decode_path, obj, (v[EOC_LEN:] if lenindef else v) def __repr__(self): return pp_console_row(next(self.pps())) @@ -3254,19 +4078,34 @@ class OctetString(Obj): yield pp -NullState = namedtuple("NullState", ( - "version", - "tag", - "expl", - "default", - "optional", - "offset", - "llen", - "vlen", - "expl_lenindef", - "lenindef", - "ber_encoded", -), **NAMEDTUPLE_KWARGS) +def agg_octet_string(evgens, decode_path, raw, writer): + """Aggregate constructed string (OctetString and its derivatives) + + :param evgens: iterator of generated events + :param decode_path: points to the string we want to decode + :param raw: slicebable (memoryview, bytearray, etc) with + the data evgens are generated on + :param writer: buffer.write where string is going to be saved + :param writer: where string is going to be saved. Must comply + with ``io.RawIOBase.write`` behaviour + + .. seealso:: :ref:`agg_octet_string` + """ + decode_path_len = len(decode_path) + for dp, obj, _ in evgens: + if dp[:decode_path_len] != decode_path: + continue + if not obj.ber_encoded: + write_full(writer, raw[ + obj.offset + obj.tlen + obj.llen: + obj.offset + obj.tlen + obj.llen + obj.vlen - + (EOC_LEN if obj.expl_lenindef else 0) + ]) + if len(dp) == decode_path_len: + break + + +NullState = namedtuple("NullState", BasicState._fields, **NAMEDTUPLE_KWARGS) class Null(Obj): @@ -3305,6 +4144,7 @@ class Null(Obj): return NullState( __version__, self.tag, + self._tag_order, self._expl, self.default, self.optional, @@ -3316,19 +4156,6 @@ class Null(Obj): self.ber_encoded, ) - def __setstate__(self, state): - super(Null, self).__setstate__(state) - self.tag = state.tag - self._expl = state.expl - self.default = state.default - self.optional = state.optional - self.offset = state.offset - self.llen = state.llen - self.vlen = state.vlen - self.expl_lenindef = state.expl_lenindef - self.lenindef = state.lenindef - self.ber_encoded = state.ber_encoded - def __eq__(self, their): if not issubclass(their.__class__, Null): return False @@ -3351,9 +4178,15 @@ class Null(Obj): ) def _encode(self): - return self.tag + len_encode(0) + return self.tag + LEN0 - def _decode(self, tlv, offset, decode_path, ctx, tag_only): + def _encode1st(self, state): + return len(self.tag) + 1, state + + def _encode2nd(self, writer, state_iter): + write_full(writer, self.tag + LEN0) + + def _decode(self, tlv, offset, decode_path, ctx, tag_only, evgen_mode): try: t, _, lv = tag_strip(tlv) except DecodeError as err: @@ -3370,7 +4203,8 @@ class Null(Obj): offset=offset, ) if tag_only: # pragma: no cover - return None + yield None + return try: l, _, v = len_decode(lv) except DecodeError as err: @@ -3393,7 +4227,7 @@ class Null(Obj): optional=self.optional, _decoded=(offset, 1, 0), ) - return obj, v + yield decode_path, obj, v def __repr__(self): return pp_console_row(next(self.pps())) @@ -3422,21 +4256,11 @@ class Null(Obj): yield pp -ObjectIdentifierState = namedtuple("ObjectIdentifierState", ( - "version", - "value", - "tag", - "expl", - "default", - "optional", - "offset", - "llen", - "vlen", - "expl_lenindef", - "lenindef", - "ber_encoded", - "defines", -), **NAMEDTUPLE_KWARGS) +ObjectIdentifierState = namedtuple( + "ObjectIdentifierState", + BasicState._fields + ("value", "defines"), + **NAMEDTUPLE_KWARGS +) class ObjectIdentifier(Obj): @@ -3483,7 +4307,9 @@ class ObjectIdentifier(Obj): tuple element is ``{OID: pyderasn.Obj()}`` dictionary, mapping between current OID value and structure applied to defined field. - :ref:`Read about DEFINED BY ` + + .. seealso:: :ref:`definedby` + :param bytes impl: override default tag with ``IMPLICIT`` one :param bytes expl: override default tag with ``EXPLICIT`` one :param default: set default value. Type same as in ``value`` @@ -3506,7 +4332,7 @@ class ObjectIdentifier(Obj): def __add__(self, their): if their.__class__ == tuple: - return self.__class__(self._value + their) + return self.__class__(self._value + array("L", their)) if isinstance(their, self.__class__): return self.__class__(self._value + their._value) raise InvalidValueType((self.__class__, tuple)) @@ -3516,10 +4342,15 @@ class ObjectIdentifier(Obj): return value._value if isinstance(value, string_types): try: - value = tuple(pureint(arc) for arc in value.split(".")) + value = array("L", (pureint(arc) for arc in value.split("."))) except ValueError: raise InvalidOID("unacceptable arcs values") if value.__class__ == tuple: + try: + value = array("L", value) + except OverflowError as err: + raise InvalidOID(repr(err)) + if value.__class__ is array: if len(value) < 2: raise InvalidOID("less than 2 arcs") first_arc = value[0] @@ -3542,8 +4373,8 @@ class ObjectIdentifier(Obj): def __getstate__(self): return ObjectIdentifierState( __version__, - self._value, self.tag, + self._tag_order, self._expl, self.default, self.optional, @@ -3553,22 +4384,13 @@ class ObjectIdentifier(Obj): self.expl_lenindef, self.lenindef, self.ber_encoded, + self._value, self.defines, ) def __setstate__(self, state): super(ObjectIdentifier, self).__setstate__(state) self._value = state.value - self.tag = state.tag - self._expl = state.expl - self.default = state.default - self.optional = state.optional - self.offset = state.offset - self.llen = state.llen - self.vlen = state.vlen - self.expl_lenindef = state.expl_lenindef - self.lenindef = state.lenindef - self.ber_encoded = state.ber_encoded self.defines = state.defines def __iter__(self): @@ -3580,15 +4402,15 @@ class ObjectIdentifier(Obj): def __hash__(self): self._assert_ready() - return hash( - self.tag + - bytes(self._expl or b"") + + return hash(b"".join(( + self.tag, + bytes(self._expl or b""), str(self._value).encode("ascii"), - ) + ))) def __eq__(self, their): if their.__class__ == tuple: - return self._value == their + return self._value == array("L", their) if not issubclass(their.__class__, ObjectIdentifier): return False return ( @@ -3618,7 +4440,7 @@ class ObjectIdentifier(Obj): optional=self.optional if optional is None else optional, ) - def _encode(self): + def _encode_octets(self): self._assert_ready() value = self._value first_value = value[1] @@ -3634,10 +4456,20 @@ class ObjectIdentifier(Obj): octets = [zero_ended_encode(first_value)] for arc in value[2:]: octets.append(zero_ended_encode(arc)) - v = b"".join(octets) + return b"".join(octets) + + def _encode(self): + v = self._encode_octets() return b"".join((self.tag, len_encode(len(v)), v)) - def _decode(self, tlv, offset, decode_path, ctx, tag_only): + def _encode1st(self, state): + l = len(self._encode_octets()) + return len(self.tag) + len_size(l) + l, state + + def _encode2nd(self, writer, state_iter): + write_full(writer, self._encode()) + + def _decode(self, tlv, offset, decode_path, ctx, tag_only, evgen_mode): try: t, _, lv = tag_strip(tlv) except DecodeError as err: @@ -3654,7 +4486,8 @@ class ObjectIdentifier(Obj): offset=offset, ) if tag_only: # pragma: no cover - return None + yield None + return try: l, llen, v = len_decode(lv) except DecodeError as err: @@ -3679,7 +4512,7 @@ class ObjectIdentifier(Obj): offset=offset, ) v, tail = v[:l], v[l:] - arcs = [] + arcs = array("L") ber_encoded = False while len(v) > 0: i = 0 @@ -3690,10 +4523,23 @@ class ObjectIdentifier(Obj): if ctx.get("bered", False): ber_encoded = True else: - raise DecodeError("non normalized arc encoding") + raise DecodeError( + "non normalized arc encoding", + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) arc = (arc << 7) | (octet & 0x7F) if octet & 0x80 == 0: - arcs.append(arc) + try: + arcs.append(arc) + except OverflowError: + raise DecodeError( + "too huge value for local unsigned long", + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) v = v[i + 1:] break i += 1 @@ -3715,7 +4561,7 @@ class ObjectIdentifier(Obj): first_arc = 2 second_arc -= 80 obj = self.__class__( - value=tuple([first_arc, second_arc] + arcs[1:]), + value=array("L", (first_arc, second_arc)) + arcs[1:], impl=self.tag, expl=self._expl, default=self.default, @@ -3724,7 +4570,7 @@ class ObjectIdentifier(Obj): ) if ber_encoded: obj.ber_encoded = True - return obj, tail + yield decode_path, obj, tail def __repr__(self): return pp_console_row(next(self.pps())) @@ -4146,8 +4992,11 @@ class IA5String(CommonString): LEN_YYMMDDHHMMSSZ = len("YYMMDDHHMMSSZ") +LEN_LEN_YYMMDDHHMMSSZ = len_encode(LEN_YYMMDDHHMMSSZ) +LEN_YYMMDDHHMMSSZ_WITH_LEN = len(LEN_LEN_YYMMDDHHMMSSZ) + LEN_YYMMDDHHMMSSZ LEN_YYYYMMDDHHMMSSDMZ = len("YYYYMMDDHHMMSSDMZ") LEN_YYYYMMDDHHMMSSZ = len("YYYYMMDDHHMMSSZ") +LEN_LEN_YYYYMMDDHHMMSSZ = len_encode(LEN_YYYYMMDDHHMMSSZ) class VisibleString(CommonString): @@ -4208,6 +5057,7 @@ class UTCTime(VisibleString): tag_default = tag_encode(23) encoding = "ascii" asn1_type_name = "UTCTime" + evgen_mode_skip_value = False def __init__( self, @@ -4334,6 +5184,7 @@ class UTCTime(VisibleString): if self.ber_encoded: value += " (%s)" % self.ber_raw return value + return None def __unicode__(self): if self.ready: @@ -4375,8 +5226,17 @@ class UTCTime(VisibleString): def _encode(self): self._assert_ready() - value = self._encode_time() - return b"".join((self.tag, len_encode(len(value)), value)) + return b"".join((self.tag, LEN_LEN_YYMMDDHHMMSSZ, self._encode_time())) + + def _encode1st(self, state): + return len(self.tag) + LEN_YYMMDDHHMMSSZ_WITH_LEN, state + + def _encode2nd(self, writer, state_iter): + self._assert_ready() + write_full(writer, self._encode()) + + def _encode_cer(self, writer): + write_full(writer, self._encode()) def todatetime(self): return self._value @@ -4539,6 +5399,22 @@ class GeneralizedTime(UTCTime): encoded += (".%06d" % value.microsecond).rstrip("0") return (encoded + "Z").encode("ascii") + def _encode(self): + self._assert_ready() + value = self._value + if value.microsecond > 0: + encoded = self._encode_time() + return b"".join((self.tag, len_encode(len(encoded)), encoded)) + return b"".join((self.tag, LEN_LEN_YYYYMMDDHHMMSSZ, self._encode_time())) + + def _encode1st(self, state): + self._assert_ready() + vlen = len(self._encode_time()) + return len(self.tag) + len_size(vlen) + vlen, state + + def _encode2nd(self, writer, state_iter): + write_full(writer, self._encode()) + class GraphicString(CommonString): __slots__ = () @@ -4573,21 +5449,11 @@ class BMPString(CommonString): asn1_type_name = "BMPString" -ChoiceState = namedtuple("ChoiceState", ( - "version", - "specs", - "value", - "tag", - "expl", - "default", - "optional", - "offset", - "llen", - "vlen", - "expl_lenindef", - "lenindef", - "ber_encoded", -), **NAMEDTUPLE_KWARGS) +ChoiceState = namedtuple( + "ChoiceState", + BasicState._fields + ("specs", "value",), + **NAMEDTUPLE_KWARGS +) class Choice(Obj): @@ -4664,6 +5530,9 @@ class Choice(Obj): self.default = default_obj if value is None: self._value = copy(default_obj._value) + if self._expl is not None: + tag_class, _, tag_num = tag_decode(self._expl) + self._tag_order = (tag_class, tag_num) def _value_sanitize(self, value): if (value.__class__ == tuple) and len(value) == 2: @@ -4692,9 +5561,8 @@ class Choice(Obj): def __getstate__(self): return ChoiceState( __version__, - self.specs, - copy(self._value), self.tag, + self._tag_order, self._expl, self.default, self.optional, @@ -4704,21 +5572,14 @@ class Choice(Obj): self.expl_lenindef, self.lenindef, self.ber_encoded, + self.specs, + copy(self._value), ) def __setstate__(self, state): super(Choice, self).__setstate__(state) self.specs = state.specs self._value = state.value - self._expl = state.expl - self.default = state.default - self.optional = state.optional - self.offset = state.offset - self.llen = state.llen - self.vlen = state.vlen - self.expl_lenindef = state.expl_lenindef - self.lenindef = state.lenindef - self.ber_encoded = state.ber_encoded def __eq__(self, their): if (their.__class__ == tuple) and len(their) == 2: @@ -4747,14 +5608,27 @@ class Choice(Obj): @property def choice(self): + """Name of the choice + """ self._assert_ready() return self._value[0] @property def value(self): + """Value of underlying choice + """ self._assert_ready() return self._value[1] + @property + def tag_order(self): + self._assert_ready() + return self._value[1].tag_order if self._tag_order is None else self._tag_order + + @property + def tag_order_cer(self): + return min(v.tag_order_cer for v in itervalues(self.specs)) + def __getitem__(self, key): if key not in self.specs: raise ObjUnknown(key) @@ -4785,7 +5659,18 @@ class Choice(Obj): self._assert_ready() return self._value[1].encode() - def _decode(self, tlv, offset, decode_path, ctx, tag_only): + def _encode1st(self, state): + self._assert_ready() + return self._value[1].encode1st(state) + + def _encode2nd(self, writer, state_iter): + self._value[1].encode2nd(writer, state_iter) + + def _encode_cer(self, writer): + self._assert_ready() + self._value[1].encode_cer(writer) + + def _decode(self, tlv, offset, decode_path, ctx, tag_only, evgen_mode): for choice, spec in iteritems(self.specs): sub_decode_path = decode_path + (choice,) try: @@ -4808,15 +5693,28 @@ class Choice(Obj): offset=offset, ) if tag_only: # pragma: no cover - return None - value, tail = spec.decode( - tlv, - offset=offset, - leavemm=True, - decode_path=sub_decode_path, - ctx=ctx, - _ctx_immutable=False, - ) + yield None + return + if evgen_mode: + for _decode_path, value, tail in spec.decode_evgen( + tlv, + offset=offset, + leavemm=True, + decode_path=sub_decode_path, + ctx=ctx, + _ctx_immutable=False, + ): + yield _decode_path, value, tail + else: + _, value, tail = next(spec.decode_evgen( + tlv, + offset=offset, + leavemm=True, + decode_path=sub_decode_path, + ctx=ctx, + _ctx_immutable=False, + _evgen_mode=False, + )) obj = self.__class__( schema=self.specs, expl=self._expl, @@ -4825,7 +5723,7 @@ class Choice(Obj): _decoded=(offset, 0, value.fulllen), ) obj._value = (choice, value) - return obj, tail + yield decode_path, obj, tail def __repr__(self): value = pp_console_row(next(self.pps())) @@ -4892,27 +5790,18 @@ class PrimitiveTypes(Choice): )) -AnyState = namedtuple("AnyState", ( - "version", - "value", - "tag", - "expl", - "optional", - "offset", - "llen", - "vlen", - "expl_lenindef", - "lenindef", - "ber_encoded", - "defined", -), **NAMEDTUPLE_KWARGS) +AnyState = namedtuple( + "AnyState", + BasicState._fields + ("value", "defined"), + **NAMEDTUPLE_KWARGS +) class Any(Obj): """``ANY`` special type >>> Any(Integer(-123)) - ANY 020185 + ANY INTEGER -123 (0X:7B) >>> a = Any(OctetString(b"hello world").encode()) ANY 040b68656c6c6f20776f726c64 >>> hexenc(bytes(a)) @@ -4932,28 +5821,47 @@ class Any(Obj): """ :param value: set the value. Either any kind of pyderasn's **ready** object, or bytes. Pay attention that - **no** validation is performed is raw binary value - is valid TLV + **no** validation is performed if raw binary value + is valid TLV, except just tag decoding :param bytes expl: override default tag with ``EXPLICIT`` one :param bool optional: is object ``OPTIONAL`` in sequence """ super(Any, self).__init__(None, expl, None, optional, _decoded) - self._value = None if value is None else self._value_sanitize(value) + if value is None: + self._value = None + else: + value = self._value_sanitize(value) + self._value = value + if self._expl is None: + if value.__class__ == binary_type: + tag_class, _, tag_num = tag_decode(tag_strip(value)[0]) + else: + tag_class, tag_num = value.tag_order + else: + tag_class, _, tag_num = tag_decode(self._expl) + self._tag_order = (tag_class, tag_num) self.defined = None def _value_sanitize(self, value): if value.__class__ == binary_type: + if len(value) == 0: + raise ValueError("Any value can not be empty") return value if isinstance(value, self.__class__): return value._value - if isinstance(value, Obj): - return value.encode() - raise InvalidValueType((self.__class__, Obj, binary_type)) + if not isinstance(value, Obj): + raise InvalidValueType((self.__class__, Obj, binary_type)) + return value @property def ready(self): return self._value is not None + @property + def tag_order(self): + self._assert_ready() + return self._tag_order + @property def bered(self): if self.expl_lenindef or self.lenindef: @@ -4965,9 +5873,10 @@ class Any(Obj): def __getstate__(self): return AnyState( __version__, - self._value, self.tag, + self._tag_order, self._expl, + None, self.optional, self.offset, self.llen, @@ -4975,28 +5884,24 @@ class Any(Obj): self.expl_lenindef, self.lenindef, self.ber_encoded, + self._value, self.defined, ) def __setstate__(self, state): super(Any, self).__setstate__(state) self._value = state.value - self.tag = state.tag - self._expl = state.expl - self.optional = state.optional - self.offset = state.offset - self.llen = state.llen - self.vlen = state.vlen - self.expl_lenindef = state.expl_lenindef - self.lenindef = state.lenindef - self.ber_encoded = state.ber_encoded self.defined = state.defined def __eq__(self, their): if their.__class__ == binary_type: - return self._value == their + if self._value.__class__ == binary_type: + return self._value == their + return self._value.encode() == their if issubclass(their.__class__, Any): - return self._value == their._value + if self.ready and their.ready: + return bytes(self) == bytes(their) + return self.ready == their.ready return False def __call__( @@ -5013,7 +5918,10 @@ class Any(Obj): def __bytes__(self): self._assert_ready() - return self._value + value = self._value + if value.__class__ == binary_type: + return value + return self._value.encode() @property def tlen(self): @@ -5021,9 +5929,34 @@ class Any(Obj): def _encode(self): self._assert_ready() - return self._value + value = self._value + if value.__class__ == binary_type: + return value + return value.encode() + + def _encode1st(self, state): + self._assert_ready() + value = self._value + if value.__class__ == binary_type: + return len(value), state + return value.encode1st(state) + + def _encode2nd(self, writer, state_iter): + value = self._value + if value.__class__ == binary_type: + write_full(writer, value) + else: + value.encode2nd(writer, state_iter) + + def _encode_cer(self, writer): + self._assert_ready() + value = self._value + if value.__class__ == binary_type: + write_full(writer, value) + else: + value.encode_cer(writer) - def _decode(self, tlv, offset, decode_path, ctx, tag_only): + def _decode(self, tlv, offset, decode_path, ctx, tag_only, evgen_mode): try: t, tlen, lv = tag_strip(tlv) except DecodeError as err: @@ -5060,14 +5993,15 @@ class Any(Obj): chunk_i += 1 tlvlen = tlen + llen + vlen + EOC_LEN obj = self.__class__( - value=tlv[:tlvlen].tobytes(), + value=None if evgen_mode else tlv[:tlvlen].tobytes(), expl=self._expl, optional=self.optional, _decoded=(offset, 0, tlvlen), ) obj.lenindef = True obj.tag = t.tobytes() - return obj, v[EOC_LEN:] + yield decode_path, obj, v[EOC_LEN:] + return except DecodeError as err: raise err.__class__( msg=err.msg, @@ -5085,24 +6019,32 @@ class Any(Obj): tlvlen = tlen + llen + l v, tail = tlv[:tlvlen], v[l:] obj = self.__class__( - value=v.tobytes(), + value=None if evgen_mode else v.tobytes(), expl=self._expl, optional=self.optional, _decoded=(offset, 0, tlvlen), ) obj.tag = t.tobytes() - return obj, tail + yield decode_path, obj, tail def __repr__(self): return pp_console_row(next(self.pps())) def pps(self, decode_path=()): + value = self._value + if value is None: + pass + elif value.__class__ == binary_type: + value = None + else: + value = repr(value) yield _pp( obj=self, asn1_type_name=self.asn1_type_name, obj_name=self.__class__.__name__, decode_path=decode_path, - blob=self._value if self.ready else None, + value=value, + blob=self._value if self._value.__class__ == binary_type else None, optional=self.optional, default=self == self.default, impl=None if self.tag == self.tag_default else tag_decode(self.tag), @@ -5132,19 +6074,6 @@ class Any(Obj): # ASN.1 constructed types ######################################################################## -def get_def_by_path(defines_by_path, sub_decode_path): - """Get define by decode path - """ - for path, define in defines_by_path: - if len(path) != len(sub_decode_path): - continue - for p1, p2 in zip(path, sub_decode_path): - if (p1 != any) and (p1 != p2): - break - else: - return define - - def abs_decode_path(decode_path, rel_path): """Create an absolute decode path from current and relative ones @@ -5170,24 +6099,26 @@ def abs_decode_path(decode_path, rel_path): return decode_path + rel_path -SequenceState = namedtuple("SequenceState", ( - "version", - "specs", - "value", - "tag", - "expl", - "default", - "optional", - "offset", - "llen", - "vlen", - "expl_lenindef", - "lenindef", - "ber_encoded", -), **NAMEDTUPLE_KWARGS) +SequenceState = namedtuple( + "SequenceState", + BasicState._fields + ("specs", "value",), + **NAMEDTUPLE_KWARGS +) + + +class SequenceEncode1stMixing(object): + def _encode1st(self, state): + state.append(0) + idx = len(state) - 1 + vlen = 0 + for v in self._values_for_encoding(): + l, _ = v.encode1st(state) + vlen += l + state[idx] = vlen + return len(self.tag) + len_size(vlen) + vlen, state -class Sequence(Obj): +class Sequence(SequenceEncode1stMixing, Obj): """``SEQUENCE`` structure type You have to make specification of sequence:: @@ -5275,6 +6206,11 @@ class Sequence(Obj): defaulted values existence validation by setting ``"allow_default_values": True`` :ref:`context ` option. + All values with DEFAULT specified are decoded atomically in + :ref:`evgen mode `. If DEFAULT value is some kind of + SEQUENCE, then it will be yielded as a single element, not + disassembled. That is required for DEFAULT existence check. + Two sequences are equal if they have equal specification (schema), implicit/explicit tagging and the same values. """ @@ -5339,9 +6275,8 @@ class Sequence(Obj): def __getstate__(self): return SequenceState( __version__, - self.specs, - {k: copy(v) for k, v in iteritems(self._value)}, self.tag, + self._tag_order, self._expl, self.default, self.optional, @@ -5351,22 +6286,14 @@ class Sequence(Obj): self.expl_lenindef, self.lenindef, self.ber_encoded, + self.specs, + {k: copy(v) for k, v in iteritems(self._value)}, ) def __setstate__(self, state): super(Sequence, self).__setstate__(state) self.specs = state.specs self._value = state.value - self.tag = state.tag - self._expl = state.expl - self.default = state.default - self.optional = state.optional - self.offset = state.offset - self.llen = state.llen - self.vlen = state.vlen - self.expl_lenindef = state.expl_lenindef - self.lenindef = state.lenindef - self.ber_encoded = state.ber_encoded def __eq__(self, their): if not isinstance(their, self.__class__): @@ -5424,22 +6351,31 @@ class Sequence(Obj): return spec.default return None - def _encoded_values(self): - raws = [] + def _values_for_encoding(self): for name, spec in iteritems(self.specs): value = self._value.get(name) if value is None: if spec.optional: continue raise ObjNotReady(name) - raws.append(value.encode()) - return raws + yield value def _encode(self): - v = b"".join(self._encoded_values()) + v = b"".join(v.encode() for v in self._values_for_encoding()) return b"".join((self.tag, len_encode(len(v)), v)) - def _decode(self, tlv, offset, decode_path, ctx, tag_only): + def _encode2nd(self, writer, state_iter): + write_full(writer, self.tag + len_encode(next(state_iter))) + for v in self._values_for_encoding(): + v.encode2nd(writer, state_iter) + + def _encode_cer(self, writer): + write_full(writer, self.tag + LENINDEF) + for v in self._values_for_encoding(): + v.encode_cer(writer) + write_full(writer, EOC) + + def _decode(self, tlv, offset, decode_path, ctx, tag_only, evgen_mode): try: t, tlen, lv = tag_strip(tlv) except DecodeError as err: @@ -5456,7 +6392,8 @@ class Sequence(Obj): offset=offset, ) if tag_only: # pragma: no cover - return None + yield None + return lenindef = False ctx_bered = ctx.get("bered", False) try: @@ -5498,23 +6435,36 @@ class Sequence(Obj): len(v) == 0 ): continue + spec_defaulted = spec.default is not None sub_decode_path = decode_path + (name,) try: - value, v_tail = spec.decode( - v, - sub_offset, - leavemm=True, - decode_path=sub_decode_path, - ctx=ctx, - _ctx_immutable=False, - ) + if evgen_mode and not spec_defaulted: + for _decode_path, value, v_tail in spec.decode_evgen( + v, + sub_offset, + leavemm=True, + decode_path=sub_decode_path, + ctx=ctx, + _ctx_immutable=False, + ): + yield _decode_path, value, v_tail + else: + _, value, v_tail = next(spec.decode_evgen( + v, + sub_offset, + leavemm=True, + decode_path=sub_decode_path, + ctx=ctx, + _ctx_immutable=False, + _evgen_mode=False, + )) except TagMismatch as err: if (len(err.decode_path) == len(decode_path) + 1) and spec.optional: continue raise defined = get_def_by_path(ctx.get("_defines", ()), sub_decode_path) - if defined is not None: + if not evgen_mode and defined is not None: defined_by, defined_spec = defined if issubclass(value.__class__, SequenceOf): for i, _value in enumerate(value): @@ -5566,31 +6516,34 @@ class Sequence(Obj): vlen += value_len sub_offset += value_len v = v_tail - if spec.default is not None and value == spec.default: - 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, - ) - values[name] = value - - spec_defines = getattr(spec, "defines", ()) - if len(spec_defines) == 0: - defines_by_path = ctx.get("defines_by_path", ()) - if len(defines_by_path) > 0: - spec_defines = get_def_by_path(defines_by_path, sub_decode_path) - if spec_defines is not None and len(spec_defines) > 0: - for rel_path, schema in spec_defines: - defined = schema.get(value, None) - if defined is not None: - ctx.setdefault("_defines", []).append(( - abs_decode_path(sub_decode_path[:-1], rel_path), - (value, defined), - )) + if spec_defaulted: + if evgen_mode: + yield sub_decode_path, value, v_tail + if value == spec.default: + 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, + ) + if not evgen_mode: + values[name] = value + spec_defines = getattr(spec, "defines", ()) + if len(spec_defines) == 0: + defines_by_path = ctx.get("defines_by_path", ()) + if len(defines_by_path) > 0: + spec_defines = get_def_by_path(defines_by_path, sub_decode_path) + if spec_defines is not None and len(spec_defines) > 0: + for rel_path, schema in spec_defines: + defined = schema.get(value, None) + if defined is not None: + ctx.setdefault("_defines", []).append(( + abs_decode_path(sub_decode_path[:-1], rel_path), + (value, defined), + )) if lenindef: if v[:EOC_LEN].tobytes() != EOC: raise DecodeError( @@ -5619,7 +6572,7 @@ class Sequence(Obj): obj._value = values obj.lenindef = lenindef obj.ber_encoded = ber_encoded - return obj, tail + yield decode_path, obj, tail def __repr__(self): value = pp_console_row(next(self.pps())) @@ -5663,7 +6616,7 @@ class Sequence(Obj): yield pp -class Set(Sequence): +class Set(Sequence, SequenceEncode1stMixing): """``SET`` structure type Its usage is identical to :py:class:`pyderasn.Sequence`. @@ -5671,8 +6624,8 @@ class Set(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 + during decode. If :ref:`bered ` context option is set, + then no error will occur. Also you can disable strict values ordering check by setting ``"allow_unordered_set": True`` :ref:`context ` option. """ @@ -5680,16 +6633,22 @@ class Set(Sequence): tag_default = tag_encode(form=TagFormConstructed, num=17) asn1_type_name = "SET" - def _encode(self): - raws = self._encoded_values() - raws.sort() - v = b"".join(raws) - return b"".join((self.tag, len_encode(len(v)), v)) + def _values_for_encoding(self): + return sorted( + super(Set, self)._values_for_encoding(), + key=attrgetter("tag_order"), + ) - def _specs_items(self): - return iteritems(self.specs) + def _encode_cer(self, writer): + write_full(writer, self.tag + LENINDEF) + for v in sorted( + super(Set, self)._values_for_encoding(), + key=attrgetter("tag_order_cer"), + ): + v.encode_cer(writer) + write_full(writer, EOC) - def _decode(self, tlv, offset, decode_path, ctx, tag_only): + def _decode(self, tlv, offset, decode_path, ctx, tag_only, evgen_mode): try: t, tlen, lv = tag_strip(tlv) except DecodeError as err: @@ -5706,7 +6665,8 @@ class Set(Sequence): offset=offset, ) if tag_only: - return None + yield None + return lenindef = False ctx_bered = ctx.get("bered", False) try: @@ -5742,12 +6702,13 @@ class Set(Sequence): 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]) + tag_order_prev = (0, 0) + _specs_items = copy(self.specs) while len(v) > 0: if lenindef and v[:EOC_LEN].tobytes() == EOC: break - for name, spec in self._specs_items(): + for name, spec in iteritems(_specs_items): sub_decode_path = decode_path + (name,) try: spec.decode( @@ -5768,16 +6729,30 @@ class Set(Sequence): decode_path=decode_path, offset=offset, ) - value, v_tail = spec.decode( - v, - sub_offset, - leavemm=True, - decode_path=sub_decode_path, - ctx=ctx, - _ctx_immutable=False, - ) + spec_defaulted = spec.default is not None + if evgen_mode and not spec_defaulted: + for _decode_path, value, v_tail in spec.decode_evgen( + v, + sub_offset, + leavemm=True, + decode_path=sub_decode_path, + ctx=ctx, + _ctx_immutable=False, + ): + yield _decode_path, value, v_tail + else: + _, value, v_tail = next(spec.decode_evgen( + v, + sub_offset, + leavemm=True, + decode_path=sub_decode_path, + ctx=ctx, + _ctx_immutable=False, + _evgen_mode=False, + )) + value_tag_order = value.tag_order value_len = value.fulllen - if value_prev.tobytes() > v[:value_len].tobytes(): + if tag_order_prev >= value_tag_order: if ctx_bered or ctx_allow_unordered_set: ber_encoded = True else: @@ -5787,22 +6762,27 @@ class Set(Sequence): 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, - ) + if spec_defaulted: + if evgen_mode: + yield sub_decode_path, value, v_tail + if 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] + del _specs_items[name] + tag_order_prev = value_tag_order sub_offset += value_len vlen += value_len v = v_tail + obj = self.__class__( schema=self.specs, impl=self.tag, @@ -5821,38 +6801,28 @@ class Set(Sequence): ) tail = v[EOC_LEN:] obj.lenindef = True - obj._value = values - if not obj.ready: - raise DecodeError( - "not all values are ready", - klass=self.__class__, - decode_path=decode_path, - offset=offset, - ) + for name, spec in iteritems(self.specs): + if name not in values and not spec.optional: + raise DecodeError( + "%s value is not ready" % name, + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) + if not evgen_mode: + obj._value = values obj.ber_encoded = ber_encoded - return obj, tail + yield decode_path, obj, tail -SequenceOfState = namedtuple("SequenceOfState", ( - "version", - "spec", - "value", - "bound_min", - "bound_max", - "tag", - "expl", - "default", - "optional", - "offset", - "llen", - "vlen", - "expl_lenindef", - "lenindef", - "ber_encoded", -), **NAMEDTUPLE_KWARGS) +SequenceOfState = namedtuple( + "SequenceOfState", + BasicState._fields + ("spec", "value", "bound_min", "bound_max"), + **NAMEDTUPLE_KWARGS +) -class SequenceOf(Obj): +class SequenceOf(SequenceEncode1stMixing, Obj): """``SEQUENCE OF`` sequence type For that kind of type you must specify the object it will carry on @@ -5878,9 +6848,21 @@ class SequenceOf(Obj): >>> ints Ints SEQUENCE OF[INTEGER 123, INTEGER 345] - Also you can initialize sequence with preinitialized values: + You can initialize sequence with preinitialized values: >>> ints = Ints([Integer(123), Integer(234)]) + + Also you can use iterator as a value: + + >>> ints = Ints(iter(Integer(i) for i in range(1000000))) + + And it won't be iterated until encoding process. Pay attention that + bounds and required schema checks are done only during the encoding + process in that case! After encode was called, then value is zeroed + back to empty list and you have to set it again. That mode is useful + mainly with CER encoding mode, where all objects from the iterable + will be streamed to the buffer, without copying all of them to + memory first. """ __slots__ = ("spec", "_bound_min", "_bound_max") tag_default = tag_encode(form=TagFormConstructed, num=16) @@ -5924,21 +6906,30 @@ class SequenceOf(Obj): self._value = copy(default_obj._value) def _value_sanitize(self, value): + iterator = False if issubclass(value.__class__, SequenceOf): value = value._value + elif hasattr(value, NEXT_ATTR_NAME): + iterator = True elif hasattr(value, "__iter__"): value = list(value) else: - raise InvalidValueType((self.__class__, iter)) - if not self._bound_min <= len(value) <= self._bound_max: - raise BoundsError(self._bound_min, len(value), self._bound_max) - for v in value: - if not isinstance(v, self.spec.__class__): - raise InvalidValueType((self.spec.__class__,)) + raise InvalidValueType((self.__class__, iter, "iterator")) + if not iterator: + if not self._bound_min <= len(value) <= self._bound_max: + raise BoundsError(self._bound_min, len(value), self._bound_max) + class_expected = self.spec.__class__ + for v in value: + if not isinstance(v, class_expected): + raise InvalidValueType((class_expected,)) return value @property def ready(self): + if hasattr(self._value, NEXT_ATTR_NAME): + return True + if self._bound_min > 0 and len(self._value) == 0: + return False return all(v.ready for v in self._value) @property @@ -5948,13 +6939,12 @@ class SequenceOf(Obj): return any(v.bered for v in self._value) def __getstate__(self): + if hasattr(self._value, NEXT_ATTR_NAME): + raise ValueError("can not pickle SequenceOf with iterator") return SequenceOfState( __version__, - self.spec, - [copy(v) for v in self._value], - self._bound_min, - self._bound_max, self.tag, + self._tag_order, self._expl, self.default, self.optional, @@ -5964,6 +6954,10 @@ class SequenceOf(Obj): self.expl_lenindef, self.lenindef, self.ber_encoded, + self.spec, + [copy(v) for v in self._value], + self._bound_min, + self._bound_max, ) def __setstate__(self, state): @@ -5972,16 +6966,6 @@ class SequenceOf(Obj): self._value = state.value self._bound_min = state.bound_min self._bound_max = state.bound_max - self.tag = state.tag - self._expl = state.expl - self.default = state.default - self.optional = state.optional - self.offset = state.offset - self.llen = state.llen - self.vlen = state.vlen - self.expl_lenindef = state.expl_lenindef - self.lenindef = state.lenindef - self.ber_encoded = state.ber_encoded def __eq__(self, their): if isinstance(their, self.__class__): @@ -6032,11 +7016,9 @@ class SequenceOf(Obj): self._value.append(value) def __iter__(self): - self._assert_ready() return iter(self._value) def __len__(self): - self._assert_ready() return len(self._value) def __setitem__(self, key, value): @@ -6047,14 +7029,83 @@ class SequenceOf(Obj): def __getitem__(self, key): return self._value[key] - def _encoded_values(self): - return [v.encode() for v in self._value] + def _values_for_encoding(self): + return iter(self._value) def _encode(self): - v = b"".join(self._encoded_values()) - return b"".join((self.tag, len_encode(len(v)), v)) + iterator = hasattr(self._value, NEXT_ATTR_NAME) + if iterator: + values = [] + values_append = values.append + class_expected = self.spec.__class__ + values_for_encoding = self._values_for_encoding() + self._value = [] + for v in values_for_encoding: + if not isinstance(v, class_expected): + raise InvalidValueType((class_expected,)) + values_append(v.encode()) + if not self._bound_min <= len(values) <= self._bound_max: + raise BoundsError(self._bound_min, len(values), self._bound_max) + value = b"".join(values) + else: + value = b"".join(v.encode() for v in self._values_for_encoding()) + return b"".join((self.tag, len_encode(len(value)), value)) + + def _encode1st(self, state): + state = super(SequenceOf, self)._encode1st(state) + if hasattr(self._value, NEXT_ATTR_NAME): + self._value = [] + return state + + def _encode2nd(self, writer, state_iter): + write_full(writer, self.tag + len_encode(next(state_iter))) + iterator = hasattr(self._value, NEXT_ATTR_NAME) + if iterator: + values_count = 0 + class_expected = self.spec.__class__ + values_for_encoding = self._values_for_encoding() + self._value = [] + for v in values_for_encoding: + if not isinstance(v, class_expected): + raise InvalidValueType((class_expected,)) + v.encode2nd(writer, state_iter) + values_count += 1 + if not self._bound_min <= values_count <= self._bound_max: + raise BoundsError(self._bound_min, values_count, self._bound_max) + else: + for v in self._values_for_encoding(): + v.encode2nd(writer, state_iter) + + def _encode_cer(self, writer): + write_full(writer, self.tag + LENINDEF) + iterator = hasattr(self._value, NEXT_ATTR_NAME) + if iterator: + class_expected = self.spec.__class__ + values_count = 0 + values_for_encoding = self._values_for_encoding() + self._value = [] + for v in values_for_encoding: + if not isinstance(v, class_expected): + raise InvalidValueType((class_expected,)) + v.encode_cer(writer) + values_count += 1 + if not self._bound_min <= values_count <= self._bound_max: + raise BoundsError(self._bound_min, values_count, self._bound_max) + else: + for v in self._values_for_encoding(): + v.encode_cer(writer) + write_full(writer, EOC) - def _decode(self, tlv, offset, decode_path, ctx, tag_only, ordering_check=False): + def _decode( + self, + tlv, + offset, + decode_path, + ctx, + tag_only, + evgen_mode, + ordering_check=False, + ): try: t, tlen, lv = tag_strip(tlv) except DecodeError as err: @@ -6071,7 +7122,8 @@ class SequenceOf(Obj): offset=offset, ) if tag_only: - return None + yield None + return lenindef = False ctx_bered = ctx.get("bered", False) try: @@ -6105,6 +7157,7 @@ class SequenceOf(Obj): vlen = 0 sub_offset = offset + tlen + llen _value = [] + _value_count = 0 ctx_allow_unordered_set = ctx.get("allow_unordered_set", False) value_prev = memoryview(v[:0]) ber_encoded = False @@ -6112,15 +7165,27 @@ class SequenceOf(Obj): 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=sub_decode_path, - ctx=ctx, - _ctx_immutable=False, - ) + sub_decode_path = decode_path + (str(_value_count),) + if evgen_mode: + for _decode_path, value, v_tail in spec.decode_evgen( + v, + sub_offset, + leavemm=True, + decode_path=sub_decode_path, + ctx=ctx, + _ctx_immutable=False, + ): + yield _decode_path, value, v_tail + else: + _, value, v_tail = next(spec.decode_evgen( + v, + sub_offset, + leavemm=True, + decode_path=sub_decode_path, + ctx=ctx, + _ctx_immutable=False, + _evgen_mode=False, + )) value_len = value.fulllen if ordering_check: if value_prev.tobytes() > v[:value_len].tobytes(): @@ -6134,13 +7199,22 @@ class SequenceOf(Obj): offset=sub_offset, ) value_prev = v[:value_len] - _value.append(value) + _value_count += 1 + if not evgen_mode: + _value.append(value) sub_offset += value_len vlen += value_len v = v_tail + if evgen_mode and not self._bound_min <= _value_count <= self._bound_max: + raise DecodeError( + msg=str(BoundsError(self._bound_min, _value_count, self._bound_max)), + klass=self.__class__, + decode_path=decode_path, + offset=offset, + ) try: obj = self.__class__( - value=_value, + value=None if evgen_mode else _value, schema=spec, bounds=(self._bound_min, self._bound_max), impl=self.tag, @@ -6167,7 +7241,7 @@ class SequenceOf(Obj): obj.lenindef = True tail = v[EOC_LEN:] obj.ber_encoded = ber_encoded - return obj, tail + yield decode_path, obj, tail def __repr__(self): return "%s[%s]" % ( @@ -6213,19 +7287,43 @@ class SetOf(SequenceOf): tag_default = tag_encode(form=TagFormConstructed, num=17) asn1_type_name = "SET OF" + def _value_sanitize(self, value): + value = super(SetOf, self)._value_sanitize(value) + if hasattr(value, NEXT_ATTR_NAME): + raise ValueError( + "SetOf does not support iterator values, as no sense in them" + ) + return value + def _encode(self): - raws = self._encoded_values() - raws.sort() - v = b"".join(raws) + v = b"".join(sorted(v.encode() for v in self._values_for_encoding())) return b"".join((self.tag, len_encode(len(v)), v)) - def _decode(self, tlv, offset, decode_path, ctx, tag_only): + def _encode2nd(self, writer, state_iter): + write_full(writer, self.tag + len_encode(next(state_iter))) + values = [] + for v in self._values_for_encoding(): + buf = BytesIO() + v.encode2nd(buf.write, state_iter) + values.append(buf.getvalue()) + values.sort() + for v in values: + write_full(writer, v) + + def _encode_cer(self, writer): + write_full(writer, self.tag + LENINDEF) + for v in sorted(encode_cer(v) for v in self._values_for_encoding()): + write_full(writer, v) + write_full(writer, EOC) + + def _decode(self, tlv, offset, decode_path, ctx, tag_only, evgen_mode): return super(SetOf, self)._decode( tlv, offset, decode_path, ctx, tag_only, + evgen_mode, ordering_check=True, ) @@ -6271,6 +7369,7 @@ def generic_decoder(): # pragma: no cover with_colours=False, with_decode_path=False, decode_path_only=(), + decode_path=(), ): def _pprint_pps(pps): for pp in pps: @@ -6302,13 +7401,399 @@ def generic_decoder(): # pragma: no cover else: for row in _pprint_pps(pp): yield row - return "\n".join(_pprint_pps(obj.pps())) + return "\n".join(_pprint_pps(obj.pps(decode_path))) return SEQUENCEOF(), pprint_any +def ascii_visualize(ba): + """Output only ASCII printable characters, like in hexdump -C + + Example output for given binary string (right part):: + + 92 2b 39 20 65 91 e6 8e 95 93 1a 58 df 02 78 ea |.+9 e......X..x.| + ^^^^^^^^^^^^^^^^ + """ + return "".join((chr(b) if 0x20 <= b <= 0x7E else ".") for b in ba) + + +def hexdump(raw): + """Generate ``hexdump -C`` like output + + Rendered example:: + + 00000000 30 80 30 80 a0 80 02 01 02 00 00 02 14 54 a5 18 |0.0..........T..| + 00000010 69 ef 8b 3f 15 fd ea ad bd 47 e0 94 81 6b 06 6a |i..?.....G...k.j| + + Result of that function is a generator of lines, where each line is + a list of columns:: + + [ + [...], + ["00000010 ", " 69", " ef", " 8b", " 3f", " 15", " fd", " ea", " ad ", + " bd", " 47", " e0", " 94", " 81", " 6b", " 06", " 6a ", + " |i..?.....G...k.j|"] + [...], + ] + """ + hexed = hexenc(raw).upper() + addr, cols = 0, ["%08x " % 0] + for i in six_xrange(0, len(hexed), 2): + if i != 0 and i // 2 % 8 == 0: + cols[-1] += " " + if i != 0 and i // 2 % 16 == 0: + cols.append(" |%s|" % ascii_visualize(bytearray(raw[addr:addr + 16]))) + yield cols + addr += 16 + cols = ["%08x " % addr] + cols.append(" " + hexed[i:i + 2]) + if len(cols) > 0: + cols.append(" |%s|" % ascii_visualize(bytearray(raw[addr:]))) + yield cols + + +def browse(raw, obj, oid_maps=()): + """Interactive browser + + :param bytes raw: binary data you decoded + :param obj: decoded :py:class:`pyderasn.Obj` + :param oid_maps: list of ``str(OID) <-> human readable string`` dictionaries. + Its human readable form is printed when OID is met + + .. note:: `urwid `__ dependency required + + This browser is an interactive terminal application for browsing + structures of your decoded ASN.1 objects. You can quit it with **q** + key. It consists of three windows: + + :tree: + View of ASN.1 elements hierarchy. You can navigate it using **Up**, + **Down**, **PageUp**, **PageDown**, **Home**, **End** keys. + **Left** key goes to constructed element above. **Plus**/**Minus** + keys collapse/uncollapse constructed elements. **Space** toggles it + :info: + window with various information about element. You can scroll it + with **h**/**l** (down, up) (**H**/**L** for triple speed) keys + :hexdump: + window with raw data hexdump and highlighted current element's + contents. It automatically focuses on element's data. You can + scroll it with **j**/**k** (down, up) (**J**/**K** for triple + speed) keys. If element has explicit tag, then it also will be + highlighted with different colour + + Window's header contains current decode path and progress bars with + position in *info* and *hexdump* windows. + + If you press **d**, then current element will be saved in the + current directory under its decode path name (adding ".0", ".1", etc + suffix if such file already exists). **D** will save it with explicit tag. + """ + from copy import deepcopy + from os.path import exists as path_exists + import urwid + + class TW(urwid.TreeWidget): + def __init__(self, state, *args, **kwargs): + self.state = state + self.scrolled = {"info": False, "hexdump": False} + super(TW, self).__init__(*args, **kwargs) + + def _get_pp(self): + pp = self.get_node().get_value() + constructed = len(pp) > 1 + return (pp if hasattr(pp, "_fields") else pp[0]), constructed + + def _state_update(self): + pp, _ = self._get_pp() + self.state["decode_path"].set_text( + ":".join(str(p) for p in pp.decode_path) + ) + lines = deepcopy(self.state["hexed"]) + + def attr_set(i, attr): + line = lines[i // 16] + idx = 1 + (i - 16 * (i // 16)) + line[idx] = (attr, line[idx]) + + if pp.expl_offset is not None: + for i in six_xrange( + pp.expl_offset, + pp.expl_offset + pp.expl_tlen + pp.expl_llen, + ): + attr_set(i, "select-expl") + for i in six_xrange(pp.offset, pp.offset + pp.tlen + pp.llen + pp.vlen): + attr_set(i, "select-value") + self.state["hexdump"]._set_body([urwid.Text(line) for line in lines]) + self.state["hexdump"].set_focus(pp.offset // 16) + self.state["hexdump"].set_focus_valign("middle") + self.state["hexdump_bar"].set_completion( + (100 * pp.offset // 16) // + len(self.state["hexdump"]._body.positions()) + ) + + lines = [ + [("header", "Name: "), pp.obj_name], + [("header", "Type: "), pp.asn1_type_name], + [("header", "Offset: "), "%d (0x%x)" % (pp.offset, pp.offset)], + [("header", "[TLV]len: "), "%d/%d/%d" % ( + pp.tlen, pp.llen, pp.vlen, + )], + [("header", "Slice: "), "[%d:%d]" % ( + pp.offset, pp.offset + pp.tlen + pp.llen + pp.vlen, + )], + ] + if pp.lenindef: + lines.append([("warning", "LENINDEF")]) + if pp.ber_encoded: + lines.append([("warning", "BER encoded")]) + if pp.bered: + lines.append([("warning", "BERed")]) + if pp.expl is not None: + lines.append([("header", "EXPLICIT")]) + klass, _, num = pp.expl + lines.append([" Tag: %s%d" % (TagClassReprs[klass], num)]) + if pp.expl_offset is not None: + lines.append([" Offset: %d" % pp.expl_offset]) + lines.append([" [TLV]len: %d/%d/%d" % ( + pp.expl_tlen, pp.expl_llen, pp.expl_vlen, + )]) + lines.append([" Slice: [%d:%d]" % ( + pp.expl_offset, + pp.expl_offset + pp.expl_tlen + pp.expl_llen + pp.expl_vlen, + )]) + if pp.impl is not None: + klass, _, num = pp.impl + lines.append([ + ("header", "IMPLICIT: "), "%s%d" % (TagClassReprs[klass], num), + ]) + if pp.optional: + lines.append(["OPTIONAL"]) + if pp.default: + lines.append(["DEFAULT"]) + if len(pp.decode_path) > 0: + ent = pp.decode_path[-1] + if isinstance(ent, DecodePathDefBy): + lines.append([""]) + value = str(ent.defined_by) + oid_name = find_oid_name( + ent.defined_by.asn1_type_name, oid_maps, value, + ) + lines.append([("header", "DEFINED BY: "), "%s" % ( + value if oid_name is None + else "%s (%s)" % (oid_name, value) + )]) + lines.append([""]) + if pp.value is not None: + lines.append([("header", "Value: "), pp.value]) + if ( + len(oid_maps) > 0 and + pp.asn1_type_name == ObjectIdentifier.asn1_type_name + ): + for oid_map in oid_maps: + oid_name = oid_map.get(pp.value) + if oid_name is not None: + lines.append([("header", "Human: "), oid_name]) + break + if pp.asn1_type_name == Integer.asn1_type_name: + lines.append([ + ("header", "Decimal: "), "%d" % int(pp.obj), + ]) + lines.append([ + ("header", "Hexadecimal: "), colonize_hex(pp.obj.tohex()), + ]) + if pp.blob.__class__ == binary_type: + blob = hexenc(pp.blob).upper() + for i in six_xrange(0, len(blob), 32): + lines.append([colonize_hex(blob[i:i + 32])]) + elif pp.blob.__class__ == tuple: + lines.append([", ".join(pp.blob)]) + self.state["info"]._set_body([urwid.Text(line) for line in lines]) + self.state["info_bar"].set_completion(0) + + def selectable(self): + if self.state["widget_current"] != self: + self.state["widget_current"] = self + self.scrolled["info"] = False + self.scrolled["hexdump"] = False + self._state_update() + return super(TW, self).selectable() + + def get_display_text(self): + pp, constructed = self._get_pp() + style = "constructed" if constructed else "" + if len(pp.decode_path) == 0: + return (style, pp.obj_name) + if pp.asn1_type_name == "EOC": + return ("eoc", "EOC") + ent = pp.decode_path[-1] + if isinstance(ent, DecodePathDefBy): + value = str(ent.defined_by) + oid_name = find_oid_name( + ent.defined_by.asn1_type_name, oid_maps, value, + ) + return ("defby", "DEFBY:" + ( + value if oid_name is None else oid_name + )) + return (style, ent) + + def _scroll(self, what, step): + self.state[what]._invalidate() + pos = self.state[what].focus_position + if not self.scrolled[what]: + self.scrolled[what] = True + pos -= 2 + pos = max(0, pos + step) + pos = min(pos, len(self.state[what]._body.positions()) - 1) + self.state[what].set_focus(pos) + self.state[what].set_focus_valign("top") + self.state[what + "_bar"].set_completion( + (100 * pos) // len(self.state[what]._body.positions()) + ) + + def keypress(self, size, key): + if key == "q": + raise urwid.ExitMainLoop() + + if key == " ": + self.expanded = not self.expanded + self.update_expanded_icon() + return None + + hexdump_steps = {"j": 1, "k": -1, "J": 5, "K": -5} + if key in hexdump_steps: + self._scroll("hexdump", hexdump_steps[key]) + return None + + info_steps = {"h": 1, "l": -1, "H": 5, "L": -5} + if key in info_steps: + self._scroll("info", info_steps[key]) + return None + + if key in ("d", "D"): + pp, _ = self._get_pp() + dp = ":".join(str(p) for p in pp.decode_path) + dp = dp.replace(" ", "_") + if dp == "": + dp = "root" + if key == "d" or pp.expl_offset is None: + data = self.state["raw"][pp.offset:( + pp.offset + pp.tlen + pp.llen + pp.vlen + )] + else: + data = self.state["raw"][pp.expl_offset:( + pp.expl_offset + pp.expl_tlen + pp.expl_llen + pp.expl_vlen + )] + ctr = 0 + + def duplicate_path(dp, ctr): + if ctr == 0: + return dp + return "%s.%d" % (dp, ctr) + + while True: + if not path_exists(duplicate_path(dp, ctr)): + break + ctr += 1 + dp = duplicate_path(dp, ctr) + with open(dp, "wb") as fd: + fd.write(data) + self.state["decode_path"].set_text( + ("warning", "Saved to: " + dp) + ) + return None + return super(TW, self).keypress(size, key) + + class PN(urwid.ParentNode): + def __init__(self, state, value, *args, **kwargs): + self.state = state + if not hasattr(value, "_fields"): + value = list(value) + super(PN, self).__init__(value, *args, **kwargs) + + def load_widget(self): + return TW(self.state, self) + + def load_child_keys(self): + value = self.get_value() + if hasattr(value, "_fields"): + return [] + return range(len(value[1:])) + + def load_child_node(self, key): + return PN( + self.state, + self.get_value()[key + 1], + parent=self, + key=key, + depth=self.get_depth() + 1, + ) + + class LabeledPG(urwid.ProgressBar): + def __init__(self, label, *args, **kwargs): + self.label = label + super(LabeledPG, self).__init__(*args, **kwargs) + + def get_text(self): + return "%s: %s" % (self.label, super(LabeledPG, self).get_text()) + + WinHexdump = urwid.ListBox([urwid.Text("")]) + WinInfo = urwid.ListBox([urwid.Text("")]) + WinDecodePath = urwid.Text("", "center") + WinInfoBar = LabeledPG("info", "pg-normal", "pg-complete") + WinHexdumpBar = LabeledPG("hexdump", "pg-normal", "pg-complete") + WinTree = urwid.TreeListBox(urwid.TreeWalker(PN( + { + "raw": raw, + "hexed": list(hexdump(raw)), + "widget_current": None, + "info": WinInfo, + "info_bar": WinInfoBar, + "hexdump": WinHexdump, + "hexdump_bar": WinHexdumpBar, + "decode_path": WinDecodePath, + }, + list(obj.pps()), + ))) + help_text = " ".join(( + "q:quit", + "space:(un)collapse", + "(pg)up/down/home/end:nav", + "jkJK:hexdump hlHL:info", + "dD:dump", + )) + urwid.MainLoop( + urwid.Frame( + urwid.Columns([ + ("weight", 1, WinTree), + ("weight", 2, urwid.Pile([ + urwid.LineBox(WinInfo), + urwid.LineBox(WinHexdump), + ])), + ]), + header=urwid.Columns([ + ("weight", 2, urwid.AttrWrap(WinDecodePath, "header")), + ("weight", 1, WinInfoBar), + ("weight", 1, WinHexdumpBar), + ]), + footer=urwid.AttrWrap(urwid.Text(help_text), "help") + ), + [ + ("header", "bold", ""), + ("constructed", "bold", ""), + ("help", "light magenta", ""), + ("warning", "light red", ""), + ("defby", "light red", ""), + ("eoc", "dark red", ""), + ("select-value", "light green", ""), + ("select-expl", "light red", ""), + ("pg-normal", "", "light blue"), + ("pg-complete", "black", "yellow"), + ], + ).run() + + def main(): # pragma: no cover import argparse - parser = argparse.ArgumentParser(description="PyDERASN ASN.1 BER/DER decoder") + parser = argparse.ArgumentParser(description="PyDERASN ASN.1 BER/CER/DER decoder") parser.add_argument( "--skip", type=int, @@ -6347,21 +7832,34 @@ def main(): # pragma: no cover help="Allow explicit tag out-of-bound", ) parser.add_argument( - "DERFile", + "--evgen", + action="store_true", + help="Turn on event generation mode", + ) + parser.add_argument( + "--browse", + action="store_true", + help="Start ASN.1 browser", + ) + parser.add_argument( + "RAWFile", type=argparse.FileType("rb"), - help="Path to DER file you want to decode", + help="Path to BER/CER/DER file you want to decode", ) args = parser.parse_args() - args.DERFile.seek(args.skip) - der = memoryview(args.DERFile.read()) - args.DERFile.close() + if PY2: + args.RAWFile.seek(args.skip) + raw = memoryview(args.RAWFile.read()) + args.RAWFile.close() + else: + raw = file_mmaped(args.RAWFile)[args.skip:] oid_maps = ( [obj_by_path(_path) for _path in (args.oids or "").split(",")] if args.oids else () ) + from functools import partial if args.schema: schema = obj_by_path(args.schema) - from functools import partial pprinter = partial(pprint, big_blobs=True) else: schema, pprinter = generic_decoder() @@ -6371,9 +7869,14 @@ def main(): # pragma: no cover } 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, + if args.browse: + obj, _ = schema().decode(raw, ctx=ctx) + browse(raw, obj, oid_maps) + from sys import exit as sys_exit + sys_exit(0) + from os import environ + pprinter = partial( + pprinter, oid_maps=oid_maps, with_colours=environ.get("NO_COLOR") is None, with_decode_path=args.print_decode_path, @@ -6381,10 +7884,17 @@ def main(): # pragma: no cover () if args.decode_path_only is None else tuple(args.decode_path_only.split(":")) ), - )) + ) + if args.evgen: + for decode_path, obj, tail in schema().decode_evgen(raw, ctx=ctx): + print(pprinter(obj, decode_path=decode_path)) + else: + obj, tail = schema().decode(raw, ctx=ctx) + print(pprinter(obj)) if tail != b"": print("\nTrailing data: %s" % hexenc(tail)) if __name__ == "__main__": + from pyderasn import * main()