I’m also curious about this.
I see marked in the package.json yet idk enough about the Logseq codebase yet to know if it’s used as the universal parser. Also there’s the mldoc parser in the Logseq Github organization (mentioned as being used here and in the main readme), yet it looks nearly abandoned.
Is there a main parser used by Logseq internals and frontend, and if so is there a place in the repo that code is centered? I’d love to contribute if I can, I’m still getting up and running with the codebase and Clojurescript :/.
Specifically, I noticed a bug in Logseq’s parsing of inline code blocks that I wanted to fix (like this GFM example yet with more backticks): can't escape 3 backticks with a 4 backtick wrapper: ```lang \n code \n ``` (trailing text)
Having a compartmentalized parser that follows a spec like GFM with Logseq-specific behavior encapsulated in an extension could make Logseq easier to develop by separating/externalizing general Markdown parsing issues while making it easier to isolate and address Logseq specific ones.
I know that could mean major refactoring, depending upon the current methodology Logseq has for parsing. It’s also a massive ask for an outsider as a first comment, really sorry about that 
. Any which way, I’d be curious to hear the latest on how Logseq is doing parsing!