[Mimedefang] strip invalid IMG tag

[Kevin A. McGrail]

> > search for <img src=(http|https)(.*)>
>
> I would *strongly* encourage you to use HTML::Parser for this task,
> and not some homegrown regexp parser.  Otherwise, you could miss
> things like:

This is just my notes.  You'll notice I did note that HTML::Parser was the
way to go.

> Why do it to text/plain parts?  Does anything actually render text/plain
> as HTML?  (Outhouse, maybe...)

Yes, I have seen anecdotal evidence that some email readers will interpret
HTML.  However, I have not found a mail reader that does but I have found
suggestions that companies like MsgTag.com believe that mail clients exist
that interpret HTML from text/plain parts.  Therefore, I am assuming a
worst-case scenario that the possibility might be there.

Regards,
KAM