drop out-of-bounds parameters in the dropInvalidMappings command

[ix0rai]

when methods change to and from static, it can occasionally leave parameters with indices that are outside of the method's parameter bounds. this means that those parameters are useless and just take up space in mappings -- they should be tossed

[IotaBread]

Since #168 they are dropped if their index is lower than the minimum. Information from the bytecode is required for the maximum

enigma/enigma/src/main/java/org/quiltmc/enigma/api/analysis/index/jar/EntryIndex.java

Lines 81 to 86 in 2e3a60d

	if (this.hasMethod(parent)) {
	// TODO: Check using max_locals from the Code attribute (JVMS§4.7.3)
	AccessFlags parentAccess = this.getMethodAccess(parent);
	int startIndex = parentAccess != null && parentAccess.isStatic() ? 0 : 1;
	return localVariableEntry.getIndex() >= startIndex;
	}

[OroArmor]

finding the number of parameters is hard since it spills over into the LVT

[OroArmor]

I just realized you can parse the descriptor and figure it out that way (plus knowing if its a static vs instance method)

[ix0rai]

am I insane? the issue I ran into implementing this is knowing if it was static based on the descriptor

[OroArmor]

I dont think there is a way to know unless you have indexed the jar. I remember aurora having this issue when working on the matcher

[IotaBread]

When matching you'd have to index the previous jar (we don't do this currently) to know if a method changed and shift its parameters accordingly. In enigma we just need to validate and drop if any parameters are invalid.
The method I mentioned in a previous comment already has access to both the method descriptor and the method access, so it'd be easy to verify there, though if we want to add local support (as an option), we need max_locals

[ix0rai]

maybe we could do it as an arg you pass to the command?