Up-to-date branch of processed files for each major version

[misode]

There would be a branch-1.X for each major version, so the structure of processed files is up-to-date with master. This is useful for third party tools that rely on mcdata.

[Arcensoth]

One complication here is that the default set of data processors that mcdata uses (which happens to be the default set included in mcgen) are kept up-to-date with the latest file structure of the server-generated files.

I think the solution might be for mcdata to retain its own copy of processors for each major version, for example in a root processors folder of this repository. This folder would contain the processors used for the particular version in any given branch.

The java command used to invoke the data generator in the server jar will also differ per version; i.e. the --cmd option will need to change around snapshot 21w39a.