Add a new have_html_escaped_body_text that match an HTML-escaped text

What does this MR do?

This solves transient failures when a text contains HTML-escapable characters such as '.

Closes #20121 (closed)