Marky

A modular and extensible Markdown parser written in TypeScript that spits out HTML available as a Deno third party module and as a ES module.

Usage

Deno

import { marky } from "https://deno.land/x/marky@v1.1.5/mod.ts";

const html = marky("**hi there**"); // => <p><strong>hi there</strong></p>

ESM

You can also use Marky anywhere where ES modules are supported by downloading and using the marky.esm.js file, and then importing it as follows:

import { marky } from "./marky.esm.js";

const html = marky("**hi there**"); // => <p><strong>hi there</strong></p>

Or if you want to use Marky in the browser and don’t want to bother downloading and hosting Marky yourself then you can import it conveniently via JSDelivr like this:

<script type="module">
import { marky } from "https://cdn.jsdelivr.net/gh/askonomm/marky@1.1.5/marky.esm.js";

document.querySelector('body').innerHTML = marky("**hi there**");
</script>

Parsers

By default Marky has the feature-set described here in the default spec section, but you can change that! You can add, remove and even change existing features however you see fit. You see, Marky is made out of a bunch of parsers as the basic building blocks of the whole thing.

If you wish to overwrite existing parsers, simply provide an array of Parser’s as the second argument to marky, like this:

function isHorizontalLineBlock(block: string): boolean {
  return block.replaceAll("\n", "").trim() === "***";
}

function horizontalLineBlock(): string {
  return `<hr>`;
}

const parsers = [{
  matcher: isHorizontalLineBlock,
  renderers: [horizontalLineBlock],
}];

const html = marky("***", parsers);

As you can see in the above example, we’re replacing the default parsers with a parser for just horizontal line blocks which will detect that a block has three asterisk characters and will then render a <hr> instead.

Each parser has an optional matcher which takes in a block as a string, and must return a boolean, in which case that parser’ renderers will be used to render the block. A parser also has an array of renderers that are used to create the final output. Each renderer takes in a block as a string, but unlike a matcher, must also return a string. You can have as many as you’d like, of course, for example:

const parsers = [{
  matcher: isHeadingBlock,
  renderers: [
    bold,
    italic,
    inlineCode,
    strikethrough,
    linkAndImage,
    headingBlock,
  ],
}];

This parser will detect if we’re dealing with a heading block, and will then run the block through several parsers.

Note: You do not have to provide a matcher for a parser, but then the parser and its renderers will only be used when no other parser was matched.

Extending default parsers

If you wish to add to the default parsers array, simply extend defaultParsers, like so:

import { defaultParsers, marky } from "./marky.esm.js";

const parsers = [{
  matcher: isMyNewBlockMatcher,
  renderers: [myNewRenderer],
}];

marky("***", defaultParsers.concat(parsers));

Selecting default parsers

You can also selectively create a list of default parsers, like so:

import {
  bold,
  headingBlock,
  isHeadingBlock,
  italic,
  marky,
} from "./marky.esm.js";

marky("***", [{
  matcher: isHeadingBlock,
  renderers: [bold, italic, headingBlock],
}]);

Available list of renderers you can import:

  • bold
  • italic
  • inlineCode
  • strikethrough
  • linkAndImage
  • emptyBlock
  • headingBlock
  • codeBlock
  • horizontalLineBlock
  • quoteBlock
  • listBlock
  • paragraphBlock

Available list of matchers you can import:

  • isEmptyBlock
  • isHeadingBlock
  • isCodeBlock
  • isHorizontalLineBlock
  • isQuoteBlock
  • isListBlock

Default spec

Bold text

Bold text is created by wrapping selected text with two asterisk characters.

There's **nothing** quite like a cold beverage on a hot summer night.

Italic text

Italic text is created by wrapping selected text with one underscore character.

There's _nothing_ quite like a cold beverage on a hot summer night.

Links can be created by wrapping the label of a link in two square brackets, followed by the link being wrapped in two parentheses.

You should totally [visit my site](https://bien.ee).

Images

Images can be created just like links, where you wrap the label (well, alt title in this case) in two square brackets which is followed by the image aadress being wrapped in two parentheses. Except, add a exclamation mark in front, which will signify that we’re dealing with an image and not with a link.

Here's a photo ![profile photo](https://somewhere.com/photo.jpg)

Inline code

Inline code text is created by wrapping selected text with one backtick character.

There's `nothing` quite like a cold beverage on a hot summer night.

Striked out text

Striked out text is created by wrapping selected text with two tilde characters.

There's ~~nothing~~ quite like a cold beverage on a hot summer night.

Horizontal line separator block

Horizontal line separator is created by having a block separated by a empty line break (just like paragraphs or code blocks) and writing three concecutive asterisk characters.

This is a paragraph.

***

And this is another paragraph separated by a horizontal line.

Paragraph blocks

Paragraphs are created by simply leaving one empty line break between text, which, technically means having two line breaks, but remember it as just one empty line between text.

Heading blocks

Headings are created by adding a octothorp (hashtag) character in front of a block of text that is separated from others by one empty line.

# This is a big title

And some paragraph goes here.

## A little smaller title

And another paragraph goes here.

As you can see, the smaller the amount of octothorp characters the bigger the title will be. You can use as many octothorps as you wish, but browsers can only recognize up to 6 of them.

Code blocks

Code blocks are created by wrapping your code with three backtick characters.

```
code goes here
```

If you want to also make sure that the HTML output would have a class associated with the programming language used in the code block, make sure to append the language name to the first occurence of backticks, like so:

```javascript
code goes here
```

Quote blocks

Quote blocks are created by prepending an arrow and space to the left of the text you want to quote.

This is a paragraph of text.

> This is a paragraph of text in a quote

Quote blocks behave like any other block, in that if you separate quote blocks by one item where there is no text (only arrow), you create new paragraphs. You can also nest quote blocks by appeding more arrows.

This is a paragraph of text.

> This is a paragraph of text in a quote
>
> This is another paragraph of text in a quote.
>
> > This is a paragraph of text in a nested quote.

List blocks

List blocks are created by prepending an asterisk character for unordered lists and a number with a dot suffix to ordered lists.

This is a paragraph.

* This is an unordered list item.
* And this is another one.

1. This is an ordered list item.
2. And this is another one.

Nested lists are also supported, which can be created with a dash character prepended to list items, like so:

* This is a list item.
- * This is a nested list item
- - 1. This is yet another level of nesting.
- * And so on.
* And so on.