improve documentation / tooling for contributions

[kevinushey]

As part of trying to update the R tree-sitter grammar, I had trouble figuring out how to replicate failing CI tests locally (as the grammar changes would require some changes to the queries bundled in this repository).

It would be helpful if there were some of the following:

1. Helper scripts to make it easier to bootstrap the requisite testing environment;
2. Documentation as to how one could prepare their local environment.

There are some helper scripts in (1), but they're not documented and seem somewhat tied to pre-requisites that only happen in CI. See e.g.

https://github.com/nvim-treesitter/nvim-treesitter/blob/master/.github/workflows/tests.yml#L33-L46

Some things that I believe would be helpful to see:

• What version of neovim should I install for running tests? How should I install it?
• What neovim plugins are required to run tests? (e.g. https://github.com/kevinushey/nvim-treesitter/blob/6bf1950ada77bd411163ca071a46bb3552c3e1fa/scripts/minimal_init.lua#L2 implies plenary.nvim is required; how should we install it?)
• While I'm looking there; should https://github.com/kevinushey/nvim-treesitter/blob/6bf1950ada77bd411163ca071a46bb3552c3e1fa/scripts/minimal_init.lua#L1 prepend to the runtime path rather than append? (so the queries etc. defined in the repository take precedence over other things that might be on runtimepath)
• How should I run a specific suite of tests? How does this differ for highlight vs. indentation tests?

I'm guessing most of this is obvious for someone familiar with the neovim ecosystem, but for those of us who are not such instructions would be incredibly helpful. Thanks for taking the time for consider!

[clason]

These are valid points, thank you for raising them!