This would be fantastic. So many plugins right now doing their own parsing, which is wasteful, or trying to operate on strings of HTML, which is infamously error-prone.

Some of my own use cases:

Process headings, figures, tables, add ids and anchors and create an outline (in eleventy-plugin-outline)
Process code blocks to highlight them, turn them into live demos, and copy code buttons etc.
Add --natural-width and --natural-height CSS properties to images

Also, if this were possible, it would also allow decoupling advanced content features from Markdown. E.g. instead of markdown-it-footnote implementing footnotes (which means they are not available outside of Markdown), it could simply transform them to a custom element (e.g. <x-footnote>), which an HTML transformation then transforms to a footnote. Same with math, citations, and many other things.

As another observation, it seems that for most use cases, the only content being transformed is content injected via content, and not the HTML of layouts.

Uh oh!

Expose HTML Transformer API #3714

Description

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions