Possible fix for multiple inline divs in a block

[dbolack-ab]

This runs a split on the html input on inline divs and then parses each singly.

This solves the issue of

<div>1</div><div>2</div><div>3</div><div>4</div><div>5</div>

producing

<div>
<p>1</p>
</div>
<div>
2
</div>
<div>
3
</div>
<div>
4
</div>
<div>
5
</div>

However the results instead are

<div>
<p>
1
</p>
</div>
<div>
<p>
2
</p>
</div>
<div>
<p>
3
</p>
</div>
<div>
<p>
4
</p>
</div>
<div>
<p>
5
</p>
</div>

If the paragraph wrappers are undesired, we either need to not run parse on the div contents or strip the paragraph wrapper.

[G-Ambatte]

The tests are failing because the final </div> is being duplicated, at least some of the time:

I haven't had a chance to look at the actual code yet to see what might be causing this duplication.

[dbolack-ab]

This does not fix nested divs

[calculuschild]

Our current behavior of parsing markdown inside block-level divs already deviates from the Commonmark standard to allow one specific case. This PR deviates even more (and suggests we should also accommodate <span> and other tags)? I could see this ballooning into a bigger and bigger thing just fighting against the commonMark standard with more and more edge cases. (This doesn't handle nested Divs, nor <div> *hello* </div> *hello*, nor <div></div> *hello* <div></div>)

At that point, it would make more sense to just overwrite the whole Markdown HTML tokenizer instead and make it into a complete extension.

Does this have an issue it's linked to?

If we have no specific high-demand user case or issue that needs this behavior, I might suggest it's better to leave this alone. We should be encouraging users to avoid raw HTML in V3 anyway.

If this is something we want to support, I would suggest a simpler approach (just split the string on <div>'s and parse everything else as markdown :

//Processes the markdown within an HTML block if it's just a class-wrapper
renderer.html = function (html) {
	if (html.match(/<\/?div.*?>/g)) {
			// Split the input HTML into `<div>` and not div
			const divBlocks = html.split(/(<\/?div.*?>)/g);

			let result = '';

			for (const block of divBlocks) {
					if(block[0] !== '<')
						result += Marked.parse(block);
					else
						result += block;
			}
			return result;
	}
	return html;
};

[dbolack-ab]

Closing pending discussion.