wishlist: EDN literals #138
Description
With edn-literals recently adopted by the CBOR WG, I'd love to see support for that in here. Implementing cri"" won't make much sense yet because CRI is not finished (when it is, I can provide the workhorse crate), but dt"" seems stable enough.
Like EDN embedded CBOR, this can be ingested without any extra processing, but producing it will need decisions (with "we don't produce it" being a viable first step). The same mechanism that would resolve #132 could be used to eventually produce them -- it'd just need generalization (because now suddenly a float or integer can either be encoded directly with a bitwidth, or the dt EDN literal).