Notation of CBOR encoded structures

[martijnharing]

When we have examples in the specification of CBOR-encoded structures they need to way to be put in the specification. This issue is to determine how to do that.

For example how do show structures like:

A26568656C6C6F6474686973626973A20103046A616E206578616D706C65

{
  "hello": "this",
  "is": {
    1: 3,
    2: "an example"
   } 
}

My proposal would be to start from CDN, as described in section 8 of https://datatracker.ietf.org/doc/html/rfc8949 and further defined in https://datatracker.ietf.org/doc/html/draft-ietf-cbor-edn-literals-27

[MasterKale]

I like CBOR previewers that break down the hex string into which characters correspond to which values, like cbor.me does:

But I wonder if we can work out a way to incorporate this while annotating example structures in a way that kinda looks like JSON instead. Here's something (cursed) to consider:

Input:

# Hex string
A26568656C6C6F6474686973626973A20103046A616E206578616D706C65

Output:

# Breakdown
[A2][65][68656C6C6F][64][74686973][62][6973][A2][01][03][04][6A][616E206578616D706C65]

# JSON-like visualization
[A2] -> map(2)  {
   [65] -> text(5) [68656C6C6F] -> "hello" : [64] -> text(4) [74686973] -> "this",
   [62] -> text(2) [6973] -> "is" : [A2] -> map(2) {
      [01] -> unsigned(1) 1: [03] -> unsigned(3) 3,
      [04] -> unsigned(4) 4: [6A] -> text(10) [616E206578616D706C65] -> "an example"
   }
}

Having individual human-readable values denote which bytes in the hex string they represent can help when debugging whether an encoder or decoder in a helper library is properly written 🤔