Fix/issue 165

[6d7a]

Closes #165

I cross-checked the encoding output in the issue using asn1.io, assuming the following ASN1 to represent the struct mentioned here:

    Issue-165 DEFINITIONS IMPLICIT TAGS ::=
    BEGIN

    Signatures ::= SEQUENCE OF MultisigPolicySignature

    MultisigPolicySignature ::= CHOICE {
        key [0] SEQUENCE {
            key [0] UTF8String
        }
    }

    END

When encoding nested SEQUENCEs, the rasn encoder did not take the custom tag into account, which in case of the nestes key-SEQUENCE and was producing a wrong tag, i.e. UNIVERSAL 0. I changed the call in the macro for encoding inner items from a simple encode to encode_with_tag. The tests run fine, and I tried to trace the different cases we could encounter through the macro process and I couldn't find any side-effects on other cases. You'll likely have a much more complete view of how the different macro calls come together, @XAMPPRocky.

Let me know if I need to do more digging.

[XAMPPRocky]

Thank you for your PR!

[orthecreedence]

Thanks @6d7a! I wasn't sure if it was an encode/decode issue...I was leaning decode, but looks like my intuition was wrong. Looks like I need to do a bit more reading of ASN1. I'll keep that asn1.io site in my back pocket from now on.

[6d7a]

No worries, thanks for the extensive documentation of the issue, that helped a lot during debugging. asn1.io is a pretty solid resource, at least as far as 5 compilations per hour get you.

[XAMPPRocky]

at least as far as 5 compilations per hour get you.

Yeah let's just a say strong motivation for getting more of the codecs working and the compiler, is to hopefully one day create a version of asn1.io that can run on GitHub pages so you can compile and test ASN.1 without the ridiculous restrictions and licences OSS put on you.