[Cbor] Invalid decimal fraction / big float?

[Laurence Lundblade <[e-mail]>]

Here’s text from section 3.4.5:

   The exponent e MUST be represented in an integer of major
   type 0 or 1, while the mantissa also can be a bignum (Section 3.4.4 <https://tools.ietf.org/html/draft-ietf-cbor-7049bis-06#section-3.4.4>).
   Contained items with other structures are invalid.

For the most part it seems clear. If the second item in the array is not an integer or a bignum the decimal fraction / big float is invalid.

However, I’ve heard comments that tags are only hints and might be optional if the meaning is clear without them. For example CWT prohibits using the tag value 1 with “exp”, “nbf” and “iat” claims.  In the case of a bignum vs an integer, the meaning is clear without a tag value 2 or 3 explicitly indicating it is a bignum.

Is this a valid decimal fraction?

C4            # tag(4)
   82         # array(2)
      01      # unsigned(1)
      42      # bytes(2)
         0102 # "\x01\x02"

Presumably this is valid, but if we did like CWT, it would be invalid.

C4               # tag(4)
   82            # array(2)
      01         # unsigned(1)
      C2         # tag(2)
         42      # bytes(2)
            0102 # "\x01\x02"

My inclination would be that they are both valid. If that is the choice, then decoders MUST handle both and that is not obvious from the text.

LL