I was thinking more that maybe some aspect of markdown was not easily expressed using a PEG (it's the same implementer behind both the parser in Pandoc and this project) .
I think code blocks don't track indentation. A code block is just a bunch of lines beginning with 4 spaces (or blank lines) and the rest (when >4 spaces) are treated as leading spaces.
I was thinking more that maybe some aspect of markdown was not easily expressed using a PEG (it's the same implementer behind both the parser in Pandoc and this project) .