The validator has_no_long_files counts the number of lines. But it might be too late if fiasko encounters a really big file: it first will try to parse it and only then run the validator. To avoid this situation, I'm suggesting to put the validator into pre-validator check category.