Re: prefer text/plain alternative over text/html

[Lloyd Zusman <ljz@asfast.com>]

Shenghuo ZHU <zsh@cs.rochester.edu> writes:

> >>>>> "Lloyd" == Lloyd Zusman <ljz@asfast.com> writes:
> 
> [...]
> 
> Lloyd> Which brings up another question: does anyone know of an
> Lloyd> existing elisp utility (maybe even something in Gnus!) which
> Lloyd> would do a quick-and-dirty-but-marginally-acceptable job of
> Lloyd> removing HTML markup tags?
> 
> `nnweb-remove-markup'. Maybe you also need `nnweb-decode-entities', or
> `nnwarchive-decode-entities' which supports decoding numbered
> entities.

Thank you very much.  After posting this question and reading your
answer, I realize that I should have asked a bit more in my original
query.  The question I really want to ask is this:

Is there a canonical Gnus method for stripping out unwanted HTML and
keeping everything from automatically being sent through the `w3'
package in the case where there is no corresponding text/plain in a
message that contains HTML?

Given that I should be using one or more of the functions you
mentioned above, is there perhaps a hook I should use for this?  Or
does this require some sort of special article treatment?  Or what?

My goal is for any such HTML to be automatically stripped and the
message not filtered through `w3', and in addition, for me to somehow
have the capability to manually cause the article to be passed again
through `w3' if I deem that to be desirable.

I can't find anything in the doc's about this.

Thanks in advance.

-- 
 Lloyd Zusman
 ljz@asfast.com