1152683 - [email] reply-to header ignored because mailbridge tries to use an array value like a singular value

Status: RESOLVED FIXED

(Firefox OS Graveyard :: Gaia::E-Mail, defect)

Description

[Raphaël Jadot]

User Agent: Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:37.0) Gecko/20100101 Firefox/37.0
Build ID: 20150403141925

Steps to reproduce:

I'm using email services that use the  “reply-to” email header for discussion lists.
I receive a message with these headers:
from = [e-mail]
reply-to = [e-mail]

I reply to the message:
 


Actual results:

The “to” field is filled with “[e-mail]”


Expected results:

The “to” field should be field with “[e-mail]”

Comment 1

[Raphaël Jadot]

I forgot to say, I could only test the email app of Firefox OS 2.0

Comment 2

[Andrew Sutherland [:asuth] (he/him)]

This is unfortunate :(.  The good news is that we do retrieve the replyTo header and parse it and try and use it.  The bad news is that there's a plurality/normalization bug; we expect the replyTo value to be a singular value when it's an array.  The original implementation was speculative as part of the initial prototyping and so didn't have a proper test case.

Unfortunately I don't see any easy workaround for this other than deleting the noreply address and making sure to have a contact that exists for the mailing list so you at least don't have to type it out every time.  This sucks and I'm very sorry about it.

NB: I did check if the standard "mailman" list options allow subscribers to decide about the reply-to thing on their own, but it does not.  It's possible other mailing list software might have that setting, but it seems like something that would be limited to just list administrators on a per-list or global basis.

Comment 3

[Andrew Sutherland [:asuth] (he/him)]

ni'ing me to pursue this in the near future, but others should feel free to take the bug if they like.  (Just be sure to add a test!)

Comment 4

[Raphaël Jadot]

(In reply to Andrew Sutherland [:asuth] from comment #2)
> Unfortunately I don't see any easy workaround for this other than deleting
> the noreply address and making sure to have a contact that exists for the
> mailing list so you at least don't have to type it out every time.  This
> sucks and I'm very sorry about it.

Thanks for reply Andrew, the workaround can't work in some cases, for example the mailing list feature of Discourse ( http://www.discourse.org/ ) which is also used by Mozilla ( 
https://discourse.mozilla-community.org ) generates a replyTo header with a message key that is never the same ( it make looks something like [e-mail] ) to identify the message to which you are replying.

Comment 5

[GitHub Autolander Bot]

Attachment: [gaia-email-libs-and-more] mcav:replyto > mozilla-b2g:master

Comment 6

[Marcus Cavanaugh [:mcav]]

Comment on attachment 8597266 [details] [review]
[gaia-email-libs-and-more] mcav:replyto > mozilla-b2g:master

Straightforward fix and test, thanks to your helpful edited bug title.

As discussed on IRC, this patch also fixes display name resolution for the reply-to address; we now respect the display name provided to Reply-To. (We do *not* fall back to the "From" display name even if reply-to doesn't include one; to do so would be confusing and likely incorrect.)

Comment 7

[Andrew Sutherland [:asuth] (he/him)]

Comment on attachment 8597266 [details] [review]
[gaia-email-libs-and-more] mcav:replyto > mozilla-b2g:master

Hooray new-style tests!  When I was imagining fixing this originally, my brain was cringing at the thought of adding more horrors to test_compose, thinking that would perhaps be the easiest route.  But this new-style test is quite delightful in both its simplicity and its lack of test_compose horrors!

Also, hooray for the nice block comment in the actual fix!

Comment 8

[Andrew Sutherland [:asuth] (he/him)]

And let's try and get this on v2.2 too.

Comment 9

[GitHub Autolander Bot]

Attachment: [gaia] mcav:replyto > mozilla-b2g:master

Comment 10

[GitHub Autolander Bot]

https://github.com/mozilla-b2g/gaia/pull/29778

Autolander could not land the pull request due to not having collaborator rights. This is possibly due to a tree closure. Please check the tree status and request checkin again once the tree is open.

Comment 11

[GitHub Autolander Bot]

Pull request has landed in master: https://github.com/mozilla-b2g/gaia/commit/c34abda6ea9ec36146eb1262c4a1a35c74e30dfa

Comment 12

[Marcus Cavanaugh [:mcav]]

Comment on attachment 8598916 [details] [review]
[gaia] mcav:replyto > mozilla-b2g:master

[Approval Request Comment]

[Bug caused by] (feature/regressing bug #): unknown

[User impact] if declined:

A "reply-to" header will be incorrectly ignored, causing users who reply to certain (typically automated) messages to send the reply to an incorrect address.

[Testing completed]: yes, automated

[Risk to taking this patch] (and alternatives if risky):

Low risk, small patch with known semantics.

[String changes made]: none.

Comment 13

[bhavana bajaj [:bajaj]]

Comment on attachment 8598916 [details] [review]
[gaia] mcav:replyto > mozilla-b2g:master

Thanks for the test here. Given , I see dogfooding users are hitting this and this would be a nice low risk improvement, uplifting it. If we have any blocking fallouts though, please back this out rather than forward fixing issues on 2.2 at this point.

Comment 14

[Ryan VanderMeulen [:RyanVM]]

v2.2: https://github.com/mozilla-b2g/gaia/commit/dc9f8d2feafe048e48c35a67a1e28fb23dfd03cb