Sanitization: clone node before importing

[swissspidy]

Context

While looking at #12159 I found another issue related to the AMP output sanitization, where some text content was completely missing.

I then realized it was because of the Sanitization_Utils::change_tag_name() method (used via Sanitization_Utils::use_semantic_heading_tags(). That util uses DomDocument::importNode(), which calls DomDocument::importNode() on a DomNodeList, which is not a simple array but actually an IteratorAggregate.

So when importing a node in a loop like that, it directly modifies the list, missing subsequent children

Summary

This PR addresses this by cloning the child nodes before importing them, making sure the DomNodeList traversal works as expected.

Relevant Technical Choices

To-do

N/A

User-facing changes

N/A

Testing Instructions

• This is a non-user-facing change and requires no QA

This PR can be tested by following these steps:

Reviews

Does this PR have a security-related impact?

No

Does this PR change what data or activity we track or use?

No

Does this PR have a legal-related impact?

No

Checklist

• This PR addresses an existing issue and I have linked this PR to it in ZenHub
• I have tested this code to the best of my abilities
• I have verified accessibility to the best of my abilities (docs)
• I have verified i18n and l10n (translation, right-to-left layout) to the best of my abilities
• This code is covered by automated tests (unit, integration, and/or e2e) to verify it works as intended (docs)
• I have added documentation where necessary
• I have added a matching Type: XYZ label to the PR

---

Fixes #

[googleforcreators-bot]

Plugin builds for e9470a3 are ready 🛎️!

• Download development build
• Download production build