edit on github

commonmark-helpers is out!

commonmark-helpers passing all the tests!

Basically, with commonmark-helpers you can find desire elements in markdown document and process them in html or plaintext format.

When I was creating this blog, I faced issue with extracting data from raw markdown also I wanted to try commonmark, because I believe It will be standard for markdown in the future. Hopefully, commonmark have rich API for traversing AST tree for your markdown documents.

Since I’m using it for several months, I can conclude, that commonmark is awesome and also that it’s API is in process of evolving. It’s not good, it’s not bad—it’s normal. I developed first version of this blog with commonmark itself, and this experience help me to understand what API helpers I need from commonmark. That was the birth of commonmark-helpers.

I was wrapping my head with these ideas for while, and then I finally implemented it today! It contains only 41 sloc, it small and very easy to useful.

Basicaly it contains three methods:

With match method you can pick any node of your markdown document: header, paragraph in simple cases, and for example first paragraph without date statement inside — in complicated cases. With a js closures you can store matching results and create something like text method:

const text = (input)=> {
  let res = '';
  match(input, (event)=> {
    res += isString(literal(event)) ? literal(event) : '';
  return res;

Yep, text is implemented with match.

Very important to notice, that this package is very simple and doesn’t contain any smart and/or complicated logic, that’s why it’s tightly coupled with commonmark API for AST tree. Check it out first.

Of course, it’s covered with tests. If you want to do something with markdown for youself try commonmark and these helpers. Patches are welcomed, If you spot any error, feel free to file an issue.

I hope these helpers will be useful for commonmark community and if so, then some of the features probably will be migrated to commonmark itself. We will see, what will happen.

Write markdown, extract metadata,
your Vladimir Starkov