test: use spawn_wait() instead of system() #31852

Problem:
Tests that need to check `nvim` CLI behavior (no RPC session) create
their own ad-hoc `system()` wrappers.

Solution:
- Use `n.spawn_wait` instead of `system()`.
- Bonus: this also improves the tests by explicitly checking for
  `stdout` or `stderr`. And if a signal is raised, `ProcStream.status`
  will reflect it.

test/functional/options/defaults_spec.lua

@@ -14,7 +14,6 @@ local api = n.api
 local command = n.command
 local clear = n.clear
 local exc_exec = n.exc_exec
-local exec_lua = n.exec_lua
 local eval = n.eval
 local eq = t.eq
 local ok = t.ok
@@ -929,17 +928,12 @@ describe('stdpath()', function()
     assert_alive() -- Check for crash. #8393
 
     -- Check that Nvim rejects invalid APPNAMEs
-    -- Call jobstart() and jobwait() in the same RPC request to reduce flakiness.
     local function test_appname(testAppname, expected_exitcode)
-      local lua_code = string.format(
-        [[
-        local child = vim.fn.jobstart({ vim.v.progpath, '--clean', '--headless', '--listen', 'x', '+qall!' }, { env = { NVIM_APPNAME = %q } })
-        return vim.fn.jobwait({ child }, %d)[1]
-      ]],
-        testAppname,
-        3000
-      )
-      eq(expected_exitcode, exec_lua(lua_code))
+      local p = n.spawn_wait({
+        args = { '--listen', 'x', '+qall!' },
+        env = { NVIM_APPNAME = testAppname },
+      })
+      eq(expected_exitcode, p.status)
     end
     -- Invalid appnames:
     test_appname('a/../b', 1)