Not on building interpreters, but on building lexers and parsers, yes. See the example for a lexical analyzer in Alex, here, combined with an intro to Happy here. I found the haskell.x and haskell.y files distributed in the darcs repos for Alex and Happy useful. You can find those here and here.