An immediately closed HTML tag does not parse properly

I’m trying to parse some user’s input HTML which happens to have an empty <div></div> tag. When parsing into XML, I get this: <div/>

And then when parsing back into HTML, the <div> is then a wrapper for the inner content.

Ex:

<div></div>
<p>hello world</p>

Parsing into XML
const xml = $.load('<div></div><p>hello world</p>', {xmlMode: true});

Parsing back into HTML:
const newHtml = xml.html({decodeEntities: true});

Gives:

<div>
<p>hello world</p>
</div>

This is an issue because of styling that might happen on the containing div, etc.

I’m not exactly sure if this is a bug or intentional. I know that <div></div> doesn’t make much sense, but it’s something I’m dealt with. To clarify; if the div has a space within it, the HTML is parsed as-expected.

[nicksheen]

fyi unicron’s repro above was obtained from the options

var content = cheerio.load('<p><strong></strong>Hello <br/>world</p>', {
                decodeEntities: false,
                xmlMode: true,
            });

This outputs xhtml content.html() === '<p><strong/>Hello <br/>world</p>'

Our difficulty with this is that we are outputting our document as html. Browsers cope with HTML containing <br/> but don’t cope with <strong/>, so we need to fix our inconsistencies. On further reflection I don’t think the issue we have experienced is with cheerio.

[fb55]

Fixed by #985