Bipullbacks of fractions and the snail lemma