Structure API Finalization

[APickledWalrus]

Description

This PR aims to make the final changes before the Structure API is "officially released" with 2.7

Changes:

• Script Event/Data methods are now marked as Experimental as they are subject to change in 2.8
  • see Internal Events API #5552
• Removed event parameters from the utility EntryData classes
  • developers should instead opt to manually set the current event in their Structure's load method (this is also to remove Bukkit dependency in these files)
• Fixed an issue where scripts could no longer have multiple options sections
• Priority is now considered for all Structure loading methods
• Fixed an issue with a deprecated reload method not replicating the same behavior
• Improved the EntryContainer get methods so that additional casting is not required for classes like List

If there any other suggested changes we want to make, let me know!

---

Target Minecraft Versions: any
Requirements: none
Related Issues: none

[TheLimeGlass]

Removal of the EntryContainer class types is just guess work. You could use TypeToken to ensure the type matches the generic. I don't see the point of unsafe type casting when it could just be returning an Object and it's up to the developer to handle proper casting, rather than Skript assuming the cast would work. This is bad practice.

Everything else is fine.

[APickledWalrus]

The main idea was to simplify how much needed to be done to just obtain the entries. e.g. it's much simpler for the developer to just be able to do: boolean myEntry = entryContainer.getOptional("myEntry", false); (especially for cases where you might have maps, but this change is certainly debatable)

I thought the constant specification of the type was more of an annoyance than anything, but I'd certainly like to hear other opinions :))

[TheLimeGlass]

The main idea was to simplify how much needed to be done to just obtain the entries. e.g. it's much simpler for the developer to just be able to do: boolean myEntry = entryContainer.getOptional("myEntry", false); (especially for cases where you might have maps, but this change is certainly debatable)

I thought the constant specification of the type was more of an annoyance than anything, but I'd certainly like to hear other opinions :))

Ya I don't agree with that. Let the developer cast it rather than assumption.

[TheLimeGlass]

My changes above and also needs to target dev/2.7 now as discussed in Discord.

TO ALL SKRIPTLANG DEVS: Please don't update this pull request to master, until Pickle changes the target branch to dev/2.7 to avoid making Pickle do revisions.

[Moderocky]

If you expect these methods to change it would probably have been better to put them in some kind of 'unsafe' class, rather than putting warnings on these.

[APickledWalrus]

These are expected to change in #5552 which is now targeting 2.8 (hence these annotations)

I'm not expecting a big impact (if any) on developers