Do you need to deal with character vectors? WIth string objects? With continued lines? With "block comments" ? With implicit continuation because you are in the middle of a [ or { definition ?
... Or is every line certain to be very simple like you show?
If every line is certain to be very simple, regexp() can find lines easily, such as
regexp(S, '^[\n%]+$', 'match', 'lineanchors')
