parsing an erroneous HTML tag

[donHenaro]

Good day to all.
In jsoup v. 1.18.2 and 1.18.3, there was a problem with parsing an erroneous HTML tag, for example: <figcaption
(if it is not closed on the right '>').

Example:
Document document = Jsoup.parser(content, "", Parser.xml Parser());

The input is an HTML fragment:

<figure>
  <img src="api/files/16f0c553-4d76-4411-84b7-5049fe01bbe0">
    <figcaption
      <span contenteditable="false">Image 3.</span>
    </figcaption>
</figure>

In previous versions (<= 1.18.1), the parser automatically fixed this, but in the current version it also cuts off the closing tag.
In the document we get:

<figure>
  <img src="api/files/16f0c553-4d76-4411-84b7-5049fe01bbe0">
    <figcaption
      <span contenteditable="false">Image 3.
    </figcaption<span>
  </img>
</figure>

[jhy]

Hi there,

I can't repro your specific output, can you double check your input and output? To make sure we're looking at the same thing.

Using your input:

<figure>
  <img src="api/files/16f0c553-4d76-4411-84b7-5049fe01bbe0">
    <figcaption
      <span contenteditable="false">Image 3.</span>
    </figcaption>
</figure>

With the HTML parser, we get:

  <figure>
   <img src="api/files/16f0c553-4d76-4411-84b7-5049fe01bbe0">
   <figcaption <span contenteditable="false">
    Image 3.
   </figcaption>
  </figure>

Note that is an element figcaption with an attribute <span. This changed in 1.18.2 in the Tokenizer; we used to create a new tag but now allow it in the attribute.

With the XML parser, we get:

<figure>
  <img src="api/files/16f0c553-4d76-4411-84b7-5049fe01bbe0">
    <figcaption _span="" contenteditable="false">Image 3.
    </figcaption>
</img></figure>

The changes to the HTML parser in 1.18.2 from the changelog:

• Follow the current HTML specification in the tokenizer to allow < as part of a tag name, instead of emitting it as a
  character node. 2230
• Similarly, allow a < as the start of an attribute name, vs creating a new element. The previous behavior was
  intended to parse closer to what we anticipated the author's intent to be, but that does not align to the spec or to
  how browsers behave. 1483

For input of HTML (not your example, but related):

<figcaption<span>Foo

We get the element figcaption<span, so the serialization is:

<figcaption<span>Foo</figcaption<span>

Which is weird, but is the HTML spec, and what current browsers do. #2230 changed to this behavior from the previous because on balance, too many issues were created by deviating from the spec. The optimist in me hopes that, because browsers will render it differently from the author's intent, those folks will review and fix their HTML.

Now I think it probably is an issue that when using the XML parser / serializer, we output the element as <figcaption<span>, it would be better as <figcaption_span>, like how we normalize the attribute <span to _span.

[donHenaro]

Thank you for your concern, the problem is small, we have already fixed the editor.
My example didn't work because of the pretty formatting.
Input HTML:

<figure>
  <img src="api/files/16f0c553-4d76-4411-84b7-5049fe01bbe0">
    <figcaption<span contenteditable="false">Image 3.</span></figcaption>
</figure>

we get:

<figure>
 <img src="api/files/16f0c553-4d76-4411-84b7-5049fe01bbe0">
  <figcaption<span contenteditable="false">
   Image 3.
  </figcaption<span>
 </img>
</figure>

Perhaps the Jsoup perceives <figcaption<span> as a whole tag and does not consider the presence of "<" inside it to be a mistake.