I am interested in adding a reader to Pandoc for the Cook markdown language. Cooklang is a concise, standardized way to format recipes. You write down the instructions for a recipe and whenever you mention a new ingredient or piece of cookware you annotate it and can specify the quantity and more. Then that document can be rendered by a variety of tools (e.g. a vscode extension I wrote) into a consistent format. This language seems like a natural addition to Pandoc and one that the Cooklang community would benefit from greatly.

I've written a parser for Cook in Haskell which would be a good start. It complies with the official Cooklang canonical tests.

Currently my parser uses Megaparsec. Do I need to refactor it to use Parsec instead or is it okay to add Megaparsec as a dependency?

Currently my parser uses String instead of Text. Is it necessary for the parser to use Text to be used by Pandoc or could I get away with the reader converting its Text input to String when parsing?

Is there anything else I should know before I start working on a PR? Does this seem like a suitable addition to Pandoc?