Destroyable - Add new sparks parameter value of "near"

[arcadeboss]

Adds a new parameter "sparks-match" to Destroyable

"sparks" must be active for "sparks-match" to have any effect

With the default value of "true" there is no change to current behavior for Destroyable sparks.

With value of "false" PGM will skip the far (+64 blocks away) logic

Purpose is to reduce the "noise" for players not actually near the action, especially for large Destroyables

Tested and verified that sound behavior changed for players and observers

[Pablete1234]

I think it needs a better name, maybe consider allowing sparks to have true, false, and 'near' that disables the global sound instead of a diff attr

[arcadeboss]

I think it needs a better name, maybe consider allowing sparks to have true, false, and 'near' that disables the global sound instead of a diff attr

I considered that, but felt wrong to change the type, but if you think it's okay, then alright.

[Pablete1234]

but felt wrong to change the type

as long as it's backwards-compatible in xml, i don't think it's an issue

[arcadeboss]

as long as it's backwards-compatible in xml, i don't think it's an issue

I have updated the code to use an Enum to enforce the 3 spark type values of "true", "false" and "near"

The code is backwards compatible, so current uses of "true" and "false" remain the same, and "near" provides the new logic of having sparks, but no far match sounds.

There is a check for valid sparks values. and a corresponding exception thrown if non-valid value provided.

[arcadeboss]

Updated from review comments. Abandoned aliasing, because couldn't make it work with parseEnum, but got everything else.

[Pablete1234]

Abandoned aliasing, because couldn't make it work with parseEnum, but got everything else.

Should be working out of the box, it's built-in already as long as you implement the aliasing interface in pgm like settings do

[arcadeboss]

Abandoned aliasing, because couldn't make it work with parseEnum, but got everything else.

Should be working out of the box, it's built-in already as long as you implement the aliasing interface in pgm like settings do

This is what I implemented. It complies, but at run time gives following error:

'sparks' attribute of 'destroyables' > 'destroyable' element @ line 2002 to 2004: Invalid format: 'false' is not a type.sparkstype (did you mean 'none'?)

See anything wrong?

public static enum SparksType implements Aliased {
NONE("false"),
NEAR("near"),
ALL("true");

private final String name;

private SparksType(String name) {
this.name = name;
}

public String getName() {
return name;
}

@NotNull
@OverRide
public Iterator iterator() {
return Iterators.forArray(name(), name);
}
}

[arcadeboss]

The first attempt at iterator() above was based on the MapPoolType class enum which has the same iterator method that you shared.

But you, also mentioned following a settings enum, so I tried implementing like SettingKey class does it, but it gets the same results at run time that first one did. It looks like this:

public static enum SparksType implements Aliased {
NONE("false"),
NEAR("near"),
ALL("true");

private final List aliases;

SparksType(String name) {
this(Collections.singletonList(name));
}

SparksType(List aliases) {
this.aliases = ImmutableList.copyOf(aliases);
}

@NotNull
@OverRide
public Iterator iterator() {
return aliases.iterator();
}
}

[Pablete1234]

Looked into why aliased enums were being rejected myself and submitted a fix into the PR myself, should be good to go now

[Pablete1234]

Also note, for backwards-compat we need to support "true", "yes", and "on" variants as those are all valid boolean parsing in pgm xml (ie: you could do sparks="off" or sparks="on" and they'd work the same as true/false)

[arcadeboss]

Wow, that's awesome, thank you for those updates!

I was poking around in the text parser, but never would have figured out those changes. Nice on the extended enums.