angezid / advanced-mark.js Goto Github PK

Hey @angezid!

I recently adopted using mark.js (more specifically, this fork of the library), and love the changes you've introduced, especially around performance. I saw you have a standing invitation from Julian to become an official maintainer on the original repo.

Would love to see these changes get merged into the upstream, so that way it can be versioned normally (right now I just have this GitHub repo's "master" branch added to my package.json). Just my two cents :)

Keep up the good work!! 🎉

Has anyone else experienced this?

Hey @angezid,

Excited to see the progress this library has made over the past few months! Curious what your thoughts are about this stale issue from the original repo regarding adding support for virtual DOMs like JSDOM.

For performance reasons, I'm looking at moving my MarkJS code to the BE, but currently can't because it still utilizes the global document object. I think maybe document could be an additional constructor prop that defaults to the global document?

What do you think?

I noticed that while this library does ship with types, the types field in package.json doesn't point to any of them. As a result, my project can't find any type definitions and I still have to alias the package back to use mark.js and use the @types/mark.js package.

The version 1 will still be supported in terms of bugs fixing - it has maximum backward compatibility with mark.js library.

Also I generated and put on GitHub Pages documentation for both versions (what it has so far).
That's all news.

A domiterator waitForIframes() method is buggy; if the context contains multiple iframes, it isn't waits for all iframes as it claims.
Also, the domiterator code, which handle iframes, is very slow.
This branch is an attempt to fix it.

This branch:

1. It waits for all context iframes and fires a single 'done' callback (the old isn't waits, as a result, fired multiple 'done' callbacks).
2. It wraps matches in a single mark element (the old wraps the same matches in a multiple mark elements).
3. It gives a significant performance boost when context is really big, no matter does it have iframes or not (the old is horribly slow, highlighting big context).
   Although for small context, due to waiting for all iframes, it can be slow.

If someone runs into problems with iframes and/or have time to test this branch, it would be nice to hear the results.

Hi @angezid, last year on mark.js repo https://github.com/julmot/mark.js/issues/458 you made a demo concerning highlighting texts under Shadow DOM.

Could you consider adding Shadow DOM support into your advanced-mark.js library?

Sample page that uses Shadow DOM: https://www.virustotal.com/

Thanks!

Hello,
Could you please add these two letters support:

őŐ
űŰ

Thank you! :)