Summary

Part of #11970

Update the Textual SPEC so that a Screen holds two keys, Key and Value, instead of simply Text.

Problem Definition

Currently our Screen struct has 3 fields: Text, Indent and Expert. In Text, we put the whole content we want to show on (ideally) one device screen. It generally takes the <key>: <value> format, e.g. From address: cosmos1abc...def. Then, on the ledger device, we do some parsing to split using the : separator:

| From address
| cosmos1ab
| c...def

Proposal

To avoid splitting, we propose to switch Screen to use 2 fields, Key and Value. This will allow less string manipulation, and seems more in line with devices implementations of Title and Content. The CBOR encoding will also be changed.