No I do not process it using lhs2tex. I just compile literate haskell code. The necessity of preprocessing kills the entire idea of having literate source. I just feed lhs files to ghc and everything works fine except code coverage.

On 30.01.2009, at 5:35, Robin Green wrote:

...

Have you tried processing the file with lhs2tex first, to unlit it? (Or is that what you are doing? It's unclear from your email.) -- Robin