Support for multi-sided signs

[PondWader]

For 1.20. This is a breaking change for when moving from 1.19 to 1.20

[extremeheat]

If it were a breaking change here, it would also propagate in mineflayer 5.0.

Usually the way changes like these are handled is just adding a new parameter or a new function to handle the new functionality, but since we have getters/setters yeah it's not future proof. Instead to preserve existing API, you can deprecate block.signText = '' and replace it with a block.setSignText(front, back) and block.getSignText(): [string, string?] on all versions.

[PondWader]

If it were a breaking change here, it would also propagate in mineflayer 5.0.

Usually the way changes like these are handled is just adding a new parameter or a new function to handle the new functionality, but since we have getters/setters yeah it's not future proof. Instead to preserve existing API, you can deprecate block.signText = '' and replace it with a block.setSignText(front, back) and block.getSignText(): [string, string?] on all versions.

As well as moving to functions, I could make the previous, to-be deprecated, APIs return data for the front of the sign which would keep it backwards compatible even when changing version to 1.20.

[PondWader]

Just doing this would avoid breaking anything when bots move to 1.20

[PondWader]

There, I've done it for bedrock too. I am a bit unsure if it's a good idea to remove blockEntity as it's even mentioned in the mineflayer docs, but if it does break a few bots I guess it doesn't matter much as it should be pretty easy for those bots to fix it

[extremeheat]

To be clear, the .blockEntity field will still exist. It will just not be overridden in the sign handling.

[rom1504]

Overall lgtm, left a few comments to resolve

[extremeheat]

Also, you're not testing 1.20.0 in the tests, you can add that to make sure this is working on 1.20

[PondWader]

For some reason when sign text is set for legacy signs, the original code pretty much ends up ignoring new lines. So setSignText('a\nb') would end up effectively being the same as setSignText('ab'), to me this feels like wrong behaviour but is this intentional?

[rom1504]

Ok looks good, merging thanks