HTML Comments interpreted as plain text in mediawiki parser

When parsing mediawiki syntax (taken directly from a wikipedia dump), I have discovered that despite the mediawiki syntax officially recognizing HTML comments, parsing mediawiki and converting to plain text gives text that still contains the HTML comments.

For example, when parsing the source of the page on anarchism, there are many instances of HTML comments when xwiki is used to translate the mediawiki to plaintext (See the end under "external links," for example).

The only viable workaround is to first convert to HTML, use an external parser to remove comments, parse this fixed HTML with xwiki, and finally convert to plain text. This seems like a lot of unnecessary (time-consuming) extra steps given that this is a well defined portion of the mediawiki syntax.