Hoping for something cleaner than regex

You may need to write an HTML to JSON helper/function, but I'm pretty sure there is something out there already doing that.

parse the HTML with DOMParser

That would only work client side though I'd imagine

That... is a good point. I take it you couldn't fool the

N/xml

module into doing your dirty work? Otherwise I'd presume it's time to resort to a library to avoid regex hell