Thread: Managesieve patch

Hi JohnDoe!

Patched my RC with your patchset (are you collaborating with Alec btw?) without problems. I'm running Dovecot 1.0 with pysieved. Everything works okay, except when trying to use "setflag". I just get a server error. I haven't tried everything else, but it seems the other functions work.. Any ideas?

Hi,

This patch is based on Alec's work but i am not actually colaborating with him. Could you send me any example of the rule you are trying to create and i will try it. I havent had any problems myself but i am testing using Dovecot Managesieve, may be there is something in the syntax pysieved doesnt like.

Ok, it's always good to colaborate

To reproduce for me:
"Match all messages"
"Mark message as" "whatever"

Would want to run ManageSieve for Dovecot, but I don't want to compile it myself just because of that.. So I'm using cmusieve, and it says that it supports a draft version of imapflags and notify (haven't tried notify myself):
LDA/Sieve - Dovecot Wiki

Please let me know if you want me to test anything.

ah you are quite right. setflag doesnt work when you select all messages. It works fine with the other 2 options :/ I will look into it.

colaboration is good but i think Alec is busy at the moment, he hasnt made a lot of updates to his patch recently.

Using cmusieve/pysieved I can't get any setflag to work (the filtering itself works if I create the rule manually, it's just when I use the web interface).

So, setting "match all/any" "Subject contains 'aa'" -> "setflag read" doesn't work here either.

Let me know if I need to test anything directly against sieve or sievec (I'm guessing that's how pysieve checks and rejects stuff?)

Please try this: in plugins/sieverules/rcube_sieve_script.php, line 44, change 'imapflags' to 'setflag'.

I found a few JS bugs just now as well so I will release an updated version in the next few days any way.

Nope sorry, still the same.

I would like to use sieve to give users the ability to build a vacation message. The MDA is ideal for this so people can choose *not* to automatically respond to spam and mailing-lists etc. Keeping this in mind, is there a way to extend this plug-in to allow users to build a rudimentary body message with a "reply" option?

I'm sorry I dont understand.