markdown.md

$\mathfrak{M}!⬇$ Markdown Ingestion

This notebook demoes feeding Clerk with markdown files. We currently make no assumption on the kind of source code passed in fenced blocks¹, we handle code as if it were clojure. Indented code blocks are treated as inert code blocks.

^:nextjournal.clerk/no-cache
(ns markdown-example
  (:require [nextjournal.clerk :as clerk]
            [nextjournal.markdown :as md]
            [nextjournal.markdown.transform :as md.transform]))

Nextjournal Markdown library is able to ingest a markdown string

(def markdown-input (slurp "https://daringfireball.net/projects/markdown/syntax.text"))


;; and parse it into a _nested clojure structure_ (an AST)


(def parsed (md/parse markdown-input))

At present, Clerk will split top level forms which are grouped together under the same cell, this is to guarantee that Clerk's dependency analysys among forms will still effectively avoid needless recomputations^[when code changes].

Parsed markdown is a tree of regular Cloure data structures² which you can manipulate with your favourite clojure functions

(def sliced (update parsed :content #(take 8 %))) ;; take just a slice 

and render back to hiccup with customisable elements.

(def renderers 
  (assoc md.transform/default-hiccup-renderers 
        :doc (partial md.transform/into-markup [:div.markdown-viewer])
        :ruler (constantly [:hr.mt-1.mb-10.border-0.w-full.h-5.bg-fuchsia-900.rounded-full])))

(def hiccup 
  (md.transform/->hiccup renderers sliced))

and finally render via Clerk's html helper.

^{::clerk/visibility :fold}
(clerk/html hiccup)

Appendix

Don't forget the closing slice 🍕 of markdown!

Footnotes

1. E.g. blocks delimited by 3 consecutive backticks. ↩

2. maps and vectors. ↩