asn1PERser 0.4.3

Parse ASN.1 schemas into Python code and encode/decode them using PER encoder

asn1PERser

This library can parse ASN.1 text schemas into Python code. ASN.1 PER (Aligned) decoder/encoder is included to use with parsed schemas.

Supported ASN.1 types and their constraints are:

ASN.1 Type	Single value	Value range	Value size	can be extended (...)	used Python contraint class
INTEGER	X	X		X	ValueRange
BOOLEAN
ENUMERATED				X	ExtensionMarker
BIT STRING	X		X	X	ValueSize
OCTET STRING	X		X	X	ValueSize
CHOICE				X	ExtensionMarker
SEQUENCE				X	ExtensionMarker
SEQUENCE OF	X		X	X	SequenceOfValueSize

Table of contents:

1. Examples
2. Decoding from string
3. Dictionary creation
4. Additional info
5. History
6. Known bugs

Examples

Following ASN.1 schema:

asn_schema = '''\
SimpleProtocol DEFINITIONS AUTOMATIC TAGS ::=
BEGIN
SimpleMessage ::= CHOICE {
    start    Start,
    stop     Stop,
    alive    Alive,
    data     Data,
    ...
}

Start ::= SEQUENCE {
    sequenceNumber    SequenceNumber,
    timestamp         UTC-Timestamp,
    srcPort           Port,
    dstPort           Port
}

Stop ::= SEQUENCE {
    sequenceNumber    SequenceNumber,
    timestamp         UTC-Timestamp,
    srcPort           Port,
    dstPort           Port
}

Alive ::= SEQUENCE {
    timestamp         UTC-Timestamp,
    ...
}

Data ::= SEQUENCE {
    sequenceNumber    SequenceNumber,
    swRelease         ENUMERATED {rel1, rel2, rel3, ...},
    macroId           BIT STRING (SIZE(20)) OPTIONAL,
    payload           Payload
}

Port ::= INTEGER (10000..65535)

SequenceNumber ::= INTEGER (0..65535)

UTC-Timestamp ::= SEQUENCE {
    seconds     INTEGER (0..4294967295),
    useconds    INTEGER (0..4294967295)
}

Payload ::= SEQUENCE (SIZE(1..5)) OF Message

Message ::= OCTET STRING (SIZE(1..4))

END
'''

can be parsed into Python code like this:

from asn1PERser import parse_asn1_schema


parse_asn1_schema(asn1_schema=asn_schema, output_folder=r'C:/my_code/')

Above code will create file 'SimpleProtocol.py' in folder (which must exist) 'C:/my_code/':

from pyasn1.type.namedtype import NamedType, NamedTypes, OptionalNamedType, DefaultedNamedType
from pyasn1.type.namedval import NamedValues
from asn1PERser.classes.data.builtin import *
from asn1PERser.classes.types.type import AdditiveNamedTypes
from asn1PERser.classes.types.constraint import MIN, MAX, NoConstraint, ExtensionMarker, SequenceOfValueSize, \
    ValueRange, SingleValue, ValueSize, ConstraintOr, ConstraintAnd


class Port(IntegerType):
    subtypeSpec = ValueRange(10000, 65535)


class SequenceNumber(IntegerType):
    subtypeSpec = ValueRange(0, 65535)


class Message(OctetStringType):
    subtypeSpec = ValueSize(1, 4)


class UTC_Timestamp(SequenceType):
    class seconds(IntegerType):
        subtypeSpec = ValueRange(0, 4294967295)

    class useconds(IntegerType):
        subtypeSpec = ValueRange(0, 4294967295)

    rootComponent = AdditiveNamedTypes(
        NamedType('seconds', seconds()),
        NamedType('useconds', useconds()),
    )
    componentType = rootComponent


class Start(SequenceType):
    rootComponent = AdditiveNamedTypes(
        NamedType('sequenceNumber', SequenceNumber()),
        NamedType('timestamp', UTC_Timestamp()),
        NamedType('srcPort', Port()),
        NamedType('dstPort', Port()),
    )
    componentType = rootComponent


class Stop(SequenceType):
    rootComponent = AdditiveNamedTypes(
        NamedType('sequenceNumber', SequenceNumber()),
        NamedType('timestamp', UTC_Timestamp()),
        NamedType('srcPort', Port()),
        NamedType('dstPort', Port()),
    )
    componentType = rootComponent


class Alive(SequenceType):
    subtypeSpec = ExtensionMarker(True)
    rootComponent = AdditiveNamedTypes(
        NamedType('timestamp', UTC_Timestamp()),
    )
    componentType = rootComponent


class Payload(SequenceOfType):
    subtypeSpec = SequenceOfValueSize(1, 5)
    componentType = Message()


class Data(SequenceType):
    class swRelease(EnumeratedType):
        subtypeSpec = ExtensionMarker(True)
        enumerationRoot = NamedValues(
            ('rel1', 0),
            ('rel2', 1),
            ('rel3', 2),
        )
        namedValues = enumerationRoot

    class macroId(BitStringType):
        subtypeSpec = ValueSize(20, 20)

    rootComponent = AdditiveNamedTypes(
        NamedType('sequenceNumber', SequenceNumber()),
        NamedType('swRelease', swRelease()),
        OptionalNamedType('macroId', macroId()),
        NamedType('payload', Payload()),
    )
    componentType = rootComponent


class SimpleMessage(ChoiceType):
    subtypeSpec = ExtensionMarker(True)
    rootComponent = AdditiveNamedTypes(
        NamedType('start', Start()),
        NamedType('stop', Stop()),
        NamedType('alive', Alive()),
        NamedType('data', Data()),
    )
    componentType = rootComponent

When schema is parsed it can be used - message can be created, encoded and decoded, using PER encoder/decoder, into bytes or Python structure:

from asn1PERser import encode, decode
from SimpleProtocol import *


'''
simple_message SimpleMessage ::= alive : {
    timestamp {
        seconds 1557528149,
        useconds 12345
    }
}
'''
utc_timestamp = UTC_Timestamp()
utc_timestamp['seconds'] = UTC_Timestamp.seconds(1557528149)
utc_timestamp['useconds'] = UTC_Timestamp.useconds(12345)

msg_alive = Alive()
msg_alive['timestamp'] = utc_timestamp

simple_message = SimpleMessage()
simple_message['alive'] = msg_alive

per_bytes = encode(asn1Spec=simple_message)
print('encoded alive bytes as hex string:')
print(per_bytes.hex())  # py2: print(binascii.hexlify(per_bytes))
print('\n')

decoded = decode(per_stream=per_bytes, asn1Spec=SimpleMessage())
print('decoded alive message structure as string...:')
print(decoded)

print('...can be accessed like dictionary:')
print(decoded['alive']['timestamp']['seconds'])
print('\n')

print('...can be transformed into dictionary:')
print(decoded.to_dict())

above will output:

encoded alive bytes as hex string:
4c5cd5fe55403039


decoded alive message structure as string...:
SimpleMessage:
 alive=Alive:
  timestamp=UTC_Timestamp:
   seconds=1557528149
   useconds=12345



...can be accessed like dictionary:
1557528149


...can be transformed into dictionary:
{'alive': OrderedDict([('timestamp', OrderedDict([('seconds', 1557528149), ('useconds', 12345)]))])}

Next message:

'''
simple_message SimpleMessage ::= start : {
    sequenceNumber    10,
    timestamp {
        seconds 1557528149,
        useconds 12345
    },
    srcPort    65533,
    dstPort    10000
}
'''

utc_timestamp = UTC_Timestamp()
utc_timestamp['seconds'] = UTC_Timestamp.seconds(1557528149)
utc_timestamp['useconds'] = UTC_Timestamp.useconds(12345)

msg_start = Start()
msg_start['sequenceNumber'] = SequenceNumber(10)
msg_start['timestamp'] = utc_timestamp
msg_start['srcPort'] = Port(65533)
msg_start['dstPort'] = Port(10000)

simple_message = SimpleMessage()
simple_message['start'] = msg_start

per_bytes = encode(asn1Spec=simple_message)
print('encoded start bytes as hex string:')
print(per_bytes.hex())  # py2: print(binascii.hexlify(per_bytes))
print('\n')

decoded = decode(per_stream=per_bytes, asn1Spec=SimpleMessage())
print('start message structure as string...:')
print(decoded)
print('...can be accessed like dictionary:')
print(decoded['start']['srcPort'])
print('\n')

print('...can be transformed into dictionary:')
print(decoded.to_dict())

above will output:

encoded start bytes as hex string:
00000ac05cd5fe55403039d8ed0000


start message structure as string...:
SimpleMessage:
 start=Start:
  sequenceNumber=10
  timestamp=UTC_Timestamp:
   seconds=1557528149
   useconds=12345

  srcPort=65533
  dstPort=10000


...can be accessed like dictionary:
65533


...can be transformed into dictionary:
{'start': OrderedDict([('sequenceNumber', 10), ('timestamp', OrderedDict([('seconds', 1557528149), ('useconds', 12345)])), ('srcPort', 65533), ('dstPort', 10000)])}

Next message:

'''
simple_message SimpleMessage ::= data : {
    sequenceNumber    55555,
    swRelease  rel2,
    macroId    '11110000111100001111'B,
    payload {
        'DEAD'H,
        'BEEF'H,
        'FEED'H,
        'AA'H,
        'BBBBBBBB'H
    }
}
'''

data_payload = Payload()
data_payload.extend([Message(hexValue='DEAD')])
data_payload.extend([Message(hexValue='BEEF')])
data_payload.extend([Message(hexValue='FEED')])
data_payload.extend([Message(hexValue='AA')])
data_payload.extend([Message(hexValue='BBBBBBBB')])

msg_data = Data()
msg_data['sequenceNumber'] = SequenceNumber(55555)
msg_data['swRelease'] = Data.swRelease('rel2')
msg_data['macroId'] = Data.macroId(binValue='11110000111100001111')
msg_data['payload'] = data_payload

simple_message = SimpleMessage()
simple_message['data'] = msg_data

per_bytes = encode(asn1Spec=simple_message)
print('encoded data bytes as hex string:')
print(per_bytes.hex())  # py2: print(binascii.hexlify(per_bytes))
print('\n')

decoded = decode(per_stream=per_bytes, asn1Spec=SimpleMessage())
print('data message structure as string...:')
print(decoded)
print('...can be accessed like dictionary:')
print(bytes(decoded['data']['payload'][0]).hex())
print('\n')

print('...can be transformed into dictionary:')
print(decoded.to_dict())

above will output:

encoded data bytes as hex string:
70d90320f0f0f880dead40beef40feed00aac0bbbbbbbb


data message structure as string...:
SimpleMessage:
 data=Data:
  sequenceNumber=55555
  swRelease=rel2
  macroId=986895
  payload=Payload:
   0xdead   0xbeef   0xfeed   0xaa   0xbbbbbbbb


...can be accessed like dictionary:
dead


...can be transformed into dictionary:
{'data': OrderedDict([('sequenceNumber', 55555), ('swRelease', 'rel2'), ('macroId', 986895), ('payload', ['dead', 'beef', 'feed', 'aa', 'bbbbbbbb'])])}

Next message:

'''
simple_message SimpleMessage ::= data : {
    sequenceNumber    256,
    swRelease  rel3,
    payload {
        'DEADBEEF'H
    }
}
'''
data_payload = Payload()
data_payload.extend([Message(hexValue='DEADBEEF')])

msg_data = Data()
msg_data['sequenceNumber'] = SequenceNumber(256)
msg_data['swRelease'] = Data.swRelease('rel3')
msg_data['payload'] = data_payload

simple_message = SimpleMessage()
simple_message['data'] = msg_data

per_bytes = encode(asn1Spec=simple_message)
print('encoded data bytes as hex string:')
print(per_bytes.hex())  # py2: print(binascii.hexlify(per_bytes))
print('\n')

decoded = decode(per_stream=per_bytes, asn1Spec=SimpleMessage())
print('data message structure as string...:')
print(decoded)
print('...can be accessed like dictionary:')
print(decoded['data']['swRelease'])
print('\n')

print('...can be transformed into dictionary:')
print(decoded.to_dict())

above will output:

encoded data bytes as hex string:
60010043deadbeef


data message structure as string...:
SimpleMessage:
 data=Data:
  sequenceNumber=256
  swRelease=rel3
  payload=Payload:
   0xdeadbeef


...can be accessed like dictionary:
rel3


...can be transformed into dictionary:
{'data': OrderedDict([('sequenceNumber', 256), ('swRelease', 'rel3'), ('payload', ['deadbeef'])])}

Decoding from string

When encoded bytes are given as string use bytearray.fromhex():

pure_string = str('70d90320f0f0f880dead40beef40feed00aac0bbbbbbbb')
real_bytes = bytearray.fromhex(pure_string)

decoded = decode(asn1Spec=SimpleMessage(), per_stream=real_bytes)
print(decoded)

above will output:

SimpleMessage:
 data=Data:
  sequenceNumber=55555
  swRelease=rel2
  macroId=986895
  payload=Payload:
   0xdead   0xbeef   0xfeed   0xaa   0xbbbbbbbb

Dictionary creation

Data above can be accessed like dictionary but it is not a dictionary. Method 'to_dict' was added to make dictionary creation simple:

msg_dict = decoded.to_dict()
print(type(msg_dict))
print(msg_dict)

output:

<class 'dict'>
{'data': OrderedDict([('sequenceNumber', 55555), ('swRelease', 'rel2'), ('macroId', 986895), ('payload', ['dead', 'beef', 'feed', 'aa', 'bbbbbbbb'])])}

To better show created structure, json.dumps() can be used:

print(json.dumps(msg_dict, indent=2))

output:

{
  "data": {
    "sequenceNumber": 55555,
    "swRelease": "rel2",
    "macroId": 986895,
    "payload": [
      "dead",
      "beef",
      "feed",
      "aa",
      "bbbbbbbb"
    ]
  }
}

Additional info

For more examples please see library tests:

• parsing tests, located in test/parsing/ folder.
• coding/encoding tests, located in test/per folder.

All above examples and tests where checked using: https://asn1.io/asn1playground/

This library inherits extensively from pyasn1 library so BER and other encoding should also work here.

Parsing may take time - when trying to parse about 2000 lines of ASN.1 it took 15-20 minutes. Tests for parsing also take time.

asn1PERser uses logging library (for now only in decoder.py module). By default logging.NullHandler() is set as library top handler so no logs are visible. If you want to enable logging call asn1perser_logging_setup with (at least) handler parameter set to something else than logging.NullHandler:

import logging
from asn1PERser import asn1perser_logging_setup

asn1perser_logging_setup(logging.StreamHandler())

Above will print logging.INFO level logs to console.

History

See CHANGES file.

Known bugs

• Defining BitStringType type with default hexValue set to empty string (''), like:

...
d15     BIT STRING DEFAULT ''H,
...

will parse to Python code but running it will cause exception:

pyasn1.error.PyAsn1Error: BitStringType.fromHexString() error: invalid literal for int() with base 16: ''

• This will parse, but sorting algorith may not sort it corretly:

    
    MySeq2 ::= MySeq1   -- ok since one level of nesting is ok

    MySeq3 ::= MySeq1   -- ok since one level of nesting is ok
    
    MySeq1 ::= SEQUENCE {
        d00     INTEGER
    }

    MySeq4 ::= MySeq2   -- NOT OK! 2nd level of nesting 

This will parse into:

MySeq4 = MySeq2


class MySeq1(SequenceType):
    rootComponent = AdditiveNamedTypes(
        NamedType('d00', IntegerType()),
    )
    componentType = rootComponent


MySeq3 = MySeq1


MySeq2 = MySeq1

so Python will not find MySeq2 - this will lead to NameError.