Hi Khaled,

I took a look at the lunamark code, and ran it on some (copied) sample. As it was much slower than I'd expected from lpeg I messed around a little and could bring down the runtime on the sample from 2 sec to 0.1 sec. Anyhow, there is room for improvement so I wonder If I should take the code, wrap it up a bit, and make a module so that we can directly process markdown in context. (Of course we then need to keep an eye on how the original develops, if mkdown develops at all.)