Encode only values for class "Universal"

[TelManuel]

I am pretty new to ASN.1, but as far as I understand the class concept, only "Universal" content can be encoded without further knowledge of the context, right?

I am having a look at LDAP ExtendedResponse for instance. I implemented it like this:

    encoder = asn1.Encoder()

    encoder.start()
    encoder.enter(asn1.Numbers.Sequence)
    encoder.write(msgId)
    encoder.enter(nr=24, cls=asn1.Classes.Application)
    encoder.write(resultCode, cls=asn1.Numbers.Enumerated)
    encoder.write(matchDn, nr=asn1.Numbers.OctetString)
    encoder.write(errorMsg, nr=asn1.Numbers.OctetString)
    encoder.write(responseName, cls=asn1.Classes.Context, nr=10)
    encoder.leave()
    encoder.leave()

    return encoder.output()

ExtendedResponse is defined in RFC4511:

ExtendedResponse ::= [APPLICATION 24] SEQUENCE {
             COMPONENTS OF LDAPResult,
             responseName     [10] LDAPOID OPTIONAL,
             responseValue    [11] OCTET STRING OPTIONAL }

LDAPOID ::= OCTET STRING -- Constrained to <numericoid>
                                 -- [RFC4512]

It defines a responseName which is "context-specific" and its tag-value is "10". This contains a LDAPOID which is encoded as OCTET STRING.

I try to write it with

    encoder.write(responseName, cls=asn1.Classes.Context, nr=10)

When I do this, the tag-value (nr) is used to derive the encoding to be used, which matches Numbers.Integer although it is OCTECT STRING. Therefore, I get the following error:

lib/ldapMessageFactory.py:192: in ldapExtendedResponse
    encoder.write(responseName, cls=asn1.Classes.Context, nr=10)
/usr/local/lib/python3.6/site-packages/asn1.py:176: in write
    value = self._encode_value(nr, value)
/usr/local/lib/python3.6/site-packages/asn1.py:265: in _encode_value
    value = self._encode_integer(value)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

value = b'1.3.6.1.4.1.1466.20036'

    @staticmethod
    def _encode_integer(value):  # type: (int) -> bytes
        """Encode an integer."""
>       if value < 0:
E       TypeError: '<' not supported between instances of 'bytes' and 'int'

/usr/local/lib/python3.6/site-packages/asn1.py:286: TypeError

So from my point of view, write has to be changed, so that it only encodes values if cls is equal to Classes.Universal.

[andrivet]

it only encodes values if cls is equal to Classes.Universal

This is fixed in version 1.4.0