Add clipboard support to //deform

[TomyLobo]

This adds a -l flag to //deform and //brush deform which, when used, will fetch blocks from the clipboard instead of the world.
Additionally, coordinate transformations that would normally be based on the selection position+size are based on the clipboard's position+size when using that mode.

Example usage:

The shape on the right is the result of copying the square shape on the left and deforming the selection with the command on screen.

This is the following brush using the same clipboard contents as before:
//brush deform -l sphere 6 x=-x;y=0;z=-z

Addresses parts of #1554

[codecov]

Codecov Report

Merging #2276 (6116518) into master (acc4b59) will increase coverage by 0.02%.
The diff coverage is 6.93%.

❗ Current head 6116518 differs from pull request most recent head 7f88a3e. Consider uploading reports for the commit 7f88a3e to get more accurate results

@@             Coverage Diff             @@
##             master   #2276      +/-   ##
===========================================
+ Coverage      8.68%   8.71%   +0.02%     
- Complexity     1041    1044       +3     
===========================================
  Files           881     883       +2     
  Lines         46291   46370      +79     
  Branches       5160    5158       -2     
===========================================
+ Hits           4022    4042      +20     
- Misses        42087   42144      +57     
- Partials        182     184       +2     

Impacted Files	Coverage Δ
...src/main/java/com/sk89q/worldedit/EditSession.java	0.77% <0.00%> (-0.01%)	⬇️
...rc/main/java/com/sk89q/worldedit/LocalSession.java	0.00% <0.00%> (ø)
...ava/com/sk89q/worldedit/command/BrushCommands.java	3.68% <0.00%> (-0.13%)	⬇️
...a/com/sk89q/worldedit/command/GeneralCommands.java	8.67% <0.00%> (-0.28%)	⬇️
...om/sk89q/worldedit/command/GenerationCommands.java	3.30% <0.00%> (ø)
...va/com/sk89q/worldedit/command/RegionCommands.java	1.20% <0.00%> (-0.05%)	⬇️
...a/com/sk89q/worldedit/function/factory/Deform.java	0.00% <0.00%> (ø)
...k89q/worldedit/math/transform/SimpleTransform.java	0.00% <0.00%> (ø)
.../regions/shape/WorldEditExpressionEnvironment.java	0.00% <0.00%> (ø)
...in/java/com/sk89q/worldedit/session/Placement.java	0.00% <0.00%> (ø)
... and 3 more

... and 4 files with indirect coverage changes

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[TomyLobo]

rebased

[TomyLobo]

rebased this on top of most of the //placement PR in order to be able to use the new Placement system for the Deform brush.
Also marked as draft because the Current code is a WIP

[TomyLobo]

Moved the (contentious and not yet rebased) //brush deform changes into a separate deform-brush-clipboard branch and rebased again.
I'll unmark this as a draft once I'm done testing if it still works.

I might rebase deform-brush-clipboard later, but that's somewhat of a failed implementation and probably should be redone from scratch anyway.

[TomyLobo]

Checks are passing, deform-brush-clipboard pushed
deform-brush-clipboard is untested.
deform-clipboard (this PR) is tested and working well, except for one issue:
for flat clipboards (i.e. clipboards which are only 1 thick on any axis) non-flat selections map some points outside the clipboard.
A workaround for the user is to set the flat axis to 0.

Two alternatives were considered:
option B: adjust only when inverting
advantage: //deform -l 1 fills the entire selection even if the clipboard only is 1 thick
disadvantage(?): any coordinate on the axis in which the clipboard only is 1 thick will yield the same raw coordinate, meaning it's impossible to go outside the box on that axis anymore, with or without -l
disadvantage: the transform's implied matrix is singular if the clipboard only is 1 thick
disadvantage: the resulting transform is technically no longer really its own inverse (though transform.inverse().inverse() == transform still holds true)

option C: add 0.5 to the region/clipboard's min/max
advantage: no singularities, continuous, no special cases
disadvantage: breaks people's existing expressions for //generate, //generatebiome, //deform and //brush deform
disadvantage: makes the relationship between block coordinates and -1..1 coordinates more complex
disadvantage: coordinates never reach -1 or 1, breaking expressions that rely on that

[TomyLobo]

Now that #2582 is merged, I rebased and force-pushed this and #2580
Will test now, but I don't expect any issues, unless the heat is getting to me more than I thought :)

[TomyLobo]

I'm done testing and enabled auto-merge.
So once someone reviews this, it's in :)

[me4502]

This looks good to me, but I'd prefer if someone else reviewed this too as I'm less familiar with the expression/deform system

[TomyLobo]

just force-pushed to change a javadoc description from input/output to source/target

[TomyLobo]

force-pushed to rename SimpleTransform to ScaleAndTranslateTransform, as suggested by @wizjany.

[TomyLobo]

I just rebased this again.
Any objections to merging it once it's done building?

[octylFractal]

I'd like to give it a review, but I can't immediately because I'm working.

[TomyLobo]

this is still done and just pending a review :)