Unexpected transformation from <em></em> tag to <em />

[Guilherme2905]

Does jsoup, when parsing XML, transform a "<em></em>" tag into "<em />"?

Was this introduced recently?

How can I see the output of an XML parser? Is there documentation?

How can I avoid this?

[jhy]

Hi,

It won't if that's the only instance of the em. See: https://try.jsoup.org/~zY6iMAg2KSVWI1XT-6sdik3rmTk

But, jsoup will infer the ability for a tag to self-close if it sees an example of such in the document. And so in the XML serialization, an empty element may be output as a self-closing tag. See e.g. https://try.jsoup.org/~MPZy4b2lpKjmsefN8xnhk39-meY

This behaviour has not changed recently. Can you check that your document has a <em /> in it? If not, something else is going on.

In #2285 I plan to make (custom) tag settings configurable, so you could override this inference.

Let me ask though -- what is the impact of this for you? XML parsers support self closing tags. And semantically both <x></x> and <x/> represent the same thing -- an empty element.

If you are feeding this eventually to an HTML parser, then why serialize as XML? Use HTML.

(Providing a code sample would be helpful here to show how you've configured jsoup and are using it.)

[Guilherme2905]

Good afternoon!

Thank you very much, the self-closing of the "em" tag really exceeded my expectations.

It doesn't exactly cause a bug for me, but it was good to be aware of the lib's standard.

I checked here the code block that imposes self-closing on the parser when encountering the occurrence