health check reports .vscode settings loaded on empty directory

[mrcjkb]

Looking at this:

rustaceanvim/lua/rustaceanvim/lsp/init.lua

Lines 43 to 48 in 10efd4d

	local function find_vscode_settings(bufname)
	local settings = {}
	local found_dirs = vim.fs.find({ '.vscode' }, { upward = true, path = vim.fs.dirname(bufname), type = 'directory' })
	if vim.tbl_isempty(found_dirs) then
	return settings
	end

If I run this:

vim.fs.find({ '.vscode' }, { upward = true, path = vim.fs.dirname(bufname), type = 'directory' })

It returns an empty table. Which then returns an empty table here:

rustaceanvim/lua/rustaceanvim/lsp/init.lua

Lines 46 to 48 in 10efd4d

	if vim.tbl_isempty(found_dirs) then
	return settings
	end

That is used here:

rustaceanvim/lua/rustaceanvim/lsp/init.lua

Lines 69 to 72 in 10efd4d

	if config.server.load_vscode_settings then
	local json_settings = find_vscode_settings(bufname)
	require('rustaceanvim.config.json').override_with_rust_analyzer_json_keys(evaluated_settings, json_settings)
	end

Which triggers this:

rustaceanvim/lua/rustaceanvim/config/json.lua

Lines 75 to 79 in 10efd4d

	function M.override_with_rust_analyzer_json_keys(tbl, json_tbl)
	M.override_with_json_keys(tbl, json_tbl, function(key)
	return vim.startswith(key, 'rust-analyzer')
	end)
	end

Which triggers this:

rustaceanvim/lua/rustaceanvim/config/json.lua

Lines 64 to 71 in 10efd4d

	function M.override_with_json_keys(tbl, json_tbl, key_predicate)
	is_json_config_loaded = true
	for json_key, value in pairs(json_tbl) do
	if not key_predicate or key_predicate(json_key) then
	override_tbl_values(tbl, json_key, value)
	end
	end
	end

Which then sets is_json_config_loaded = true, even though it was an empty table because there was no JSON.

That then triggers the "loaded vscode json" message in the healthcheck. So I don't believe that's relevant in this case, unless the fact that it things an (empty) vscode file was loaded makes it not inject the check config, but I haven't checked if that's the case.

Originally posted by @JeanMertz in #652 (reply in thread)