laravel-dev-tools VS neotest

In addition to that I use and somewhat maintain a blade language server https://github.com/haringsrob/laravel-dev-tools that gives me some useful blade autocomplete as well.

Already shared some earlier developments I did by making a blade specific LSP: https://github.com/haringsrob/laravel-dev-tools

You can give it a try with the instructions here: https://github.com/haringsrob/laravel-dev-tools

My first neovim plugin project was to write an adapter (runner) for neotest, for a language that I use regularly and which was missing from the existing adapters.

There's overseer.nvim to run all sorts of things and neotest to run tests. In general, you can check awesome-neovim or TWiN to look for plugins.

Here's some things that may help: - https://github.com/mfussenegger/nvim-dap + dap ui - https://github.com/nvim-neotest/neotest

I use neotest to run tests (with pytest).

For testing I use https://github.com/nvim-neotest/neotest which gives a nice outline.

For those who don't know it, it's an adapter for neotest, a Neovim plugin that

You might look at how Neotest handles this, or even see if you can use/contribute to that project directly. It already has support for a fair number of languages at varying levels of maturity.

I am writing Python these days, and LSP + DAP + neotest give me 90% of what PyCharm would have given me. I don't know how well this works with Java, the latter two did not exist when I was still writing Java. The way I did it was use Neovim with jdt-ls to write code and IntelliJ for debugging. Debugging does not require switching between keyboard and mouse, so it was OK to me.

I'm sure that we could figure out a way to install rocks without luarocks, but it's still going to be an additional step with possible complications. Personally, I would rather just embed the library within my plugin like neotest does with xml2lua. The marginal cost of a slightly larger install is more than paid for by an easier install process and a guarantee that it works.