Sieve to strip html from multipart messages

[David Pottage]

On 2019-06-10 20:39, @lbutlr via dovecot wrote:
> Is is possible (I’m sure it is, so how) to strip multipart messages
> that are HTML and text of the HTML portion and leave the bare text
> message without also stripping other parts (like images or attached
> files).
> 
> Or also to take messages that are only HTML and strip the HTML and
> replace it with a plain text version?
> 
> I used to do this a long time ago with procmail and lynx, but it was
> never reliable since procmail doesn’t really understand MIME.

It should be possible by piping the mail through a script, using the 
sieve-extprogram plugin [1]. You would then need to find or write a 
script to strip out the HTML. For example if you are fluent in Perl then 
the Email::MIME module would be a good starting point, or if you used to 
do it with procmail, perhaps the script you used with it could be 
adapted to suit.

If you want to strip html from *ALL* emails, then it would probably make 
more sense to put the hooks into your MTA config. See this [2[ stack 
overflow post.

1: https://wiki2.dovecot.org/Pigeonhole/Sieve/Plugins/Extprograms

2: 
https://serverfault.com/questions/506894/how-to-route-email-to-a-script

-- 
David Pottage