#pragma page-filename DOC/versions/11960323
== How do I preserve HTML formatting while filtering other content? ==

If you want to filter content with Mailman's content filtering, but preserve HTML message formatting, there are a few things you must do. You must ensure that MIME type text/html is not removed. The normal way to do this is to put 
`text/html` in pass_mime_types along with your other settings. Adding this to the distributed defaults results in

{{{
multipart/mixed
multipart/alternative
text/plain
text/html
}}}

however, since there are other multipart sub-types that are commonly used such as `multipart/related`, a better option is

{{{
multipart
text/plain
text/html
}}}

The above will allow `text/html` and `text/plain` parts to remain in the messages after the initial filtering. You may also want to allow those parts to remain if they are sub-parts of an attached forwarded message `message/rfc822` part, and you may want to allow PGP signatures `application/pgp-signature` resulting in pass_mime_types containing

{{{
multipart
message/rfc822
text/plain
text/html
application/pgp-signature
}}}

The above will allow `text/html` parts to remain after initial filtering, but in order that they pass the remaining steps unchanged, you must set

{{{
Should Mailman collapse multipart/alternative to its first part content?
(Edit collapse_alternatives)
}}}

to `No` so that a `text/html` alternative part is not removed in favor of a `text/plain` alternative, and

{{{
Should Mailman convert text/html parts to plain text?
This conversion happens after MIME attachments have been stripped.
(Edit convert_html_to_plaintext)
}}}

to `No` for the obvious reason.

Also, The list's `Non-digest options -> scrub_nondigest` setting must be `No` or the HTML part(s) will be stored aside and replaced in delivered messages by brief notes with the URL of the removed part.