Bug Description

As reported in a support forum topic, if a theme like Koji declares the Content-Type using the HTML4 method like this:

<meta http-equiv="content-type" content="<?php bloginfo( 'html_type' ); ?>" charset="<?php bloginfo( 'charset' ); ?>" />

Then the result the AMP plugin generating:

<meta content="text/html; charset=UTF-8" charset="UTF-8" http-equiv="Content-Type">

Somehow the validating sanitizer is leaking the http-equiv and content attributes here, and this causes an error in the validator:

This is even though the sanitizer is reporting the former as being invalid:

Even though it is accepted, it is not getting sanitized. So that's confusing.

The AMP plugin should transparently convert any such meta Content-Type tag into the HTML meta charset that AMP requires:

<meta charset=utf-8>

See also #2970. This would have avoided an error in adding support for Twenty Twenty (#3342 (comment)) which was fixed in Twenty Twenty via WordPress/twentytwenty#640.

Here is a workaround plugin for the issue: https://gist.github.com/westonruter/d18fd37fb9a58a26ae375a0cb9570831

There should not be any need to show a validation error as this should automatically get converted. If the site is using an encoding other than UTF-8, then technically the encoding should be getting converted to UTF-8 before being parsed. See:

This would fix #855.

Expected Behaviour

An AMP-valid HTML5 meta charset should only be output.

Steps to reproduce

1. Install and activate Koji theme.
2. Activate Standard or Transitional mode.
3. See the the AMP Validator report a validation error for http-equiv attribute.
4. See the admin bar report a validation error for the same.

Do not alter or remove anything below. The following sections will be managed by moderators only.

Acceptance criteria

• The http-equiv attribute should not be leaked to the frontend.
• Convert HTML4 meta[http-equiv] to HTML5 meta[charset].
• (Nice to have) Convert encoding of input documents to UTF-8.

Implementation brief

QA testing instructions

Demo

Changelog entry