Decoded panics (and maybe revert strings?) should have an appropriate field in interpretations

[haltman-at]

Currently, when printing out the decoded result of a revert, panics often get special handling so we can print out a human-readable interpretation of what the panic means (using the panicString function from debug-utils). However, checking for this case is still a pain.

Instead, decoded panics should get an appropriate interpretation (in interpretations), and the ReturndataDecodingInspector should make use of this interpretation. The panicString function can be moved inside Codec, and called from ReturndataDecodingInspector. This would save some repeated annoying manual special cases.

Edit: Arguably, this should apply to ordinary revert strings as well! However it's more important for panics.

Environment

• Truffle version (truffle version): 5.8.4