markdown-it-image-figures

Render images occurring by itself in a paragraph as a figure with support for figcaptions.

Usage no npm install needed!

<script type="module">
  import markdownItImageFigures from 'https://cdn.skypack.dev/markdown-it-image-figures';
</script>

README

Markdown IT Image Figures

NPM version

Render images occurring by itself in a paragraph as <figure><img ...></figure>, similar to pandoc's implicit figures.

This module is a fork from markdown-it-implicit-figures in which I wanted to introduce new features and make sure this was up to what the standard is today.

Example input:

text with ![](img.png)

![](fig.png)

works with links too:

[![](fig.png)](page.html)

Output:

<p>text with <img src="img.png" alt=""></p>
<figure><img src="fig.png" alt=""></figure>
<p>works with links too:</p>
<figure><a href="page.html"><img src="fig.png" alt=""></a></figure>

Install

$ npm i markdown-it-image-figures

Usage

const md = require('markdown-it')();
const implicitFigures = require('markdown-it-image-figures');

md.use(implicitFigures);

const src = 'text with ![](img.png)\n\n![](fig.png)\n\nanother paragraph';
const res = md.render(src);

console.log(res);

/*
<p>text with <img src="img.png" alt=""></p>
<figure><img src="fig.png" alt=""></figure>
<p>another paragraph</p>
*/

Options

  • dataType: Set dataType to true to declare the data-type being wrapped, e.g.: <figure data-type="image">. This can be useful for applying a special styling for different kind of figures.

  • figcaption: Set figcaption to true to use the title as a <figcaption> block after the image. E.g.: ![This is an alt](fig.png "This is a caption") renders to

<figure>
    <img src="fig.png" alt="This is an alt">
    <figcaption>This is a caption</figcaption>
</figure>
  • tabindex: Set tabindex to true to add a tabindex property to each figure, beginning at tabindex="1" and incrementing for each figure encountered. Could be used with this css-trick, which expands figures upon mouse-over.

  • link: Put a link around the image if there is none yet.

  • copyAttrs: Copy attributes matching (RegExp or string) copyAttrs to figure element.

  • lazy: Applies the loading attribute as lazy.

  • removeSrc: Removes the source from the image and saves it on data-src.

Code like ![alt](fig.png) renders to:

<figure>
    <img alt="alt" src="fig.png" loading="lazy">
</figure>

You can override it for a single image with something like ![alt](fig.png){loading=eager} which will generate the following markup:

<figure>
    <img alt="alt" src="fig.png" loading="eager">
</figure>
  • classes: Adds the classes to the list of classes the image might have.

  • async: Adds the attribute decoding="async" to all images. As with lazy you should be able to undo this for singular images ![alt](fig.png){decoding=auto}

Web performance recommended settings

Recommended settings for web performance is as follows

{
  lazy: true,
  async: true
}

Which will add loading="lazy" and decoding="async" to all images. This can be changed per image as explained above so you can opt out for a image at the top if you'd like. This will work great for the majority of the browsers.

However, if you need to broad your browser support and ensure that old browsers get lazy loaded images, you should probably use this setting:

md.use(implicitFigures, {
  lazy: true,
  removeSrc: true,
  async: true,
  classes: 'lazy'
});

const src = '![alt](fig.png)';
const res = md.render(src);

console.log(res);
/*
<figure>
    <img alt="alt" data-src="fig.png" class="lazy" loading="lazy" decoding="async">
</figure>
*/

Then you need to load something like Lozad.js and some script like this. You might want to customise the class on the attribute classes which get added to the img (for easy selector).