Issue Summary

Broken anchor links when importing from Cloud.
This is reproducible on Data Center: (yes) / (no)

Steps to Reproduce

1. In Confluence Cloud, create a page with an anchor macro and anchor link.
2. Still on Cloud, export the space where that page was created.
3. Import that file into Confluence Server/DC

Expected Results

The anchor links should work properly.

Actual Results

he anchor links are formatted as a web link and do not work as anchor links anymore.
Also, instead of being in the "advanced" portion of the link it is now treated as a hyperlink as per image below:

This seems to be caused by the fact that Confluence Cloud creates the anchor link as an href, differently of what Confluence on premise expects.

Here's the storage format on the Cloud instance:

<p> <a href="#title1">anchor</a>  </p>
<h3><ac:structured-macro ac:name="anchor" ac:schema-version="1" ac:local-id="b1c3b7ab-53bb-4bf1-be29-71ed8b55d128" ac:macro-id="2e7929ffcee1c65caf7613d743bb24ac"><ac:parameter ac:name="">title1</ac:parameter></ac:structured-macro>This is a title that will be anchored</h3>
<p>and some other portion of the text is here.</p>
<p />

And this is the storage format when we create the anchor on premise:

<p><ac:link ac:anchor="anchor-1"><ac:plain-text-link-body><![CDATA[ anchor ]]></ac:plain-text-link-body></ac:link></p>
<h3><ac:structured-macro ac:name="anchor" ac:schema-version="1" ac:macro-id="5858d849-ed36-4223-a8cd-4c04c6b727a6"><ac:parameter ac:name="">anchor-1</ac:parameter></ac:structured-macro>This is a title that will be anchored</h3>
<p><br /></p>
<p><br /></p>

So, basically, Confluence on premise should be able to properly identify an anchor link created on Cloud that was imported on it.

Workaround

If on the link edit we remove the http/s portion, it comes back to being an anchor link.