schema-companion.nvim

Forked from the original repository someone-stole-my-name/yaml-companion.nvim and expanded for a bit more modularity to work with multiple language servers like yaml-language-server and helm-ls at the same time as well as automatic Kubernetes CRD detection. I have been happily using the plugin with some caveats, so I wanted to refactor it a bit to match my current mostly Kubernetes working environment.

Currently in the dogfooding stage with matching all the resources, but the following features are available.

Features

• Ability to add any kind of matcher, that can detect schema based on content.
• Kubernetes resources can be matched utilizing the repository yannh/kubernetes-json-schema.
• Kubernetes CRD definitions can be matched utilizing the repository datreeio/CRDs-catalog.
• Change matcher variables on the fly, like the Kubernetes version. Do not be stuck with whatever yaml-language-server has hardcoded at the given time.
• Select one from multiple matchers for the current buffer to not have any collisions in the yaml-language-server.

Installation

lazy.nvim

return {
  "cenk1cenk2/schema-companion.nvim",
  dependencies = {
    { "nvim-lua/plenary.nvim" },
    { "nvim-telescope/telescope.nvim" },
  },
  config = function()
    require("schema-companion").setup({
      -- if you have telescope you can register the extension
      enable_telescope = true,
      matchers = {
        -- add your matchers
        require("schema-companion.matchers.kubernetes").setup({ version = "master" }),
      },
    })
  end,
}

Configuration

Plugin has to be configured once, and the language servers can be added by extending the LSP configuration.

Setup

The default plugin configuration for the setup function is as below.

require("schema-companion").setup({
  log_level = vim.log.levels.INFO,
  enable_telescope = false,
  matchers = {},
  schemas = {},
}

Language Server Configuration

You can automatically extend your configuration of the yaml-language-server or helm-ls with the following configuration.

require("lspconfig").yamlls.setup(require("schema-companion").setup_client({
  -- your yaml language server configuration
}))

Manual Schemas

You can add custom schemas that can be activated manually through the telescope picker.

schemas = {
  {
    name = "Kubernetes master",
    uri = "https://raw.githubusercontent.com/yannh/kubernetes-json-schema/master/master-standalone-strict/all.json",
  },
  {
    name = "Kubernetes v1.30",
    uri = "https://raw.githubusercontent.com/yannh/kubernetes-json-schema/master/v1.30.3-standalone-strict/all.json",
  },
}

Matchers

Adding custom matchers are as easy as defining a function that detects and handles a schema.

Usage

Schema Picker

You can map the telescope picker to any keybinding of your choice.

require("telescope").extensions.yaml_schema.select_schema()

If there are multiple matches for the buffer, you can select the schema manually from the ones that matches.

require("telescope").extensions.yaml_schema.select_from_matching_schemas()

Current Schema

local schema = require("schema-companion").get_buf_schema(vim.api.nvim_get_current_buf())

This can be further utilized in lualine as follows.

-- your lualine configuration
require("lualine").setup({
  sections = {
    lualine_c = {
      {
        function()
          return ("%s"):format(require("schema-companion.context").get_buffer_schema(0).name)
        end,
        cond = function()
          return package.loaded["schema-companion"]
        end,
      },
    },
  },
})