A note on Arzela–Ascoli's lemma in almost periodic problems