They didn't have to, several of the lines matched existing fragments.

Also, the classical corpus is so small that humans still beat the computers for authorship analysis.