Check that no line from any of the input files was lost.
This is not perfect for several reasons.
It doesn't check that the content goes to the desired location.
It doesn't check that sections are created as necessary.
It doesn't support whitespace normalization that the parsing code does.
But it's a good start.