Tiletypes GUI

[tatoyoda600]

Adds a GUI for the tiletypes command.
Requires the tiletypes interface from DFHack/dfhack#4645 to function

Additionally, I made some self-contained widget stuff for this, which could be cleaned up a bit and extracted to widgets.lua or some other file, if they seem useful:

• makeInlineButtonLabelText: Able to take in 2 widgets.ButtonLabelSpec, align them, and convert them into a single widgets.ButtonLabelSpec (Used in this GUI for the buttons in the "More Options" window)
• BoxSelection: Allows for selecting a box of tiles, has support for both cursor and mouse, lets you change what tiles the selection draws onto the screen, and supports the new DimensionsTooltip
• SelectDialog: Like the gui.materials popup but without the restriction to materials, allowing you to specify whatever list of items/categories you want (ALlows nesting categories as deep as you want)

[tatoyoda600]

Taken out of WIP, as the changes from DFHack/dfhack#4645 allow this to function as intended

[myk002]

this works? I never tested this particular format

[tatoyoda600]

Yup, I used tile_color=true to have the tile_overrides ascii characters still be colored like in the ascii view. Otherwise they would always be the standard black and white.
Of course ideally I would be using actual sprites instead of colored characters, but it's what I was able to put together.

[myk002]

sorry, I meant having a combination single- and double- index array. I've tested just the color (no array) and a double-indexed array. I never tried non-table elements in the array.

[tatoyoda600]

Ah, yeah, it didn't used to be possible until I added in the support for different combinations of tables and non-tables in DFHack/dfhack#4622 / DFHack/dfhack#4624, I specifically made sure that that combination was possible cause I knew it would be useful to me :p

[myk002]

also needs docs in docs/gui/tiletypes.rst

[tatoyoda600]

Here's the spritesheet the script now tries to access, since it can't be included as part of this PR due to submodules

[myk002]

I ran out of time at SelectDialog; will finish review later, but I wanted to submit what I had so far for feedback/discussion

[tatoyoda600]

Had to go back to clamping the keyboard cursor in bounds because I realized that guidm.getCursorPos() doesn't give consistent results on all out of bounds situations.
Since guidm.getCursorPos() only ever checks if .x >= 0, the keyboard cursor leaving the map to the left causes different results than leaving up, right, or down, which looks quite janky.

[myk002]

Looks good. I'm ok with merging this now if you are. We can start getting feedback and update as necessary.