Here we attempt to get the text out of a comment.
Once we have the markup as text, we can manipulate and parse it.
Due to the tricks we're using, the comment's text value will have
an extraneous <!--
at the end. No worries,
we can get rid of that later with text.slice(0, -4)
.