345580 - Problem decoding quoted-printable question mark in subject

Status: RESOLVED FIXED

(Core :: Networking, defect)

Description

[Christian Joergensen]

User-Agent:       Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8) Gecko/20060110 Debian/1.5.dfsg-4 Firefox/1.5
Build Identifier: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8) Gecko/20060110 Debian/1.5.dfsg-4 Firefox/1.5

When a subject contains only "Subject: =?ISO-8859-1?Q???=" Mozilla Thunderbird doesn't decode it and shows up as "=?ISO-8859-1?Q???=". The encoding is however legal with regards to RFC2045:

6.7  2)   (Literal representation) Octets with decimal values of
          33 through 60 inclusive, and 62 through 126, inclusive,
          MAY be represented as the US-ASCII characters which
          correspond to those octets (EXCLAMATION POINT through
          LESS THAN, and GREATER THAN through TILDE,
          respectively).

Thank you,
Christian

Reproducible: Always

Steps to Reproduce:
1. Create an .eml file containing:

Return-Path: <[e-mail]>
X-Original-To: [e-mail]
Delivered-To: [e-mail]
Message-ID: <[e-mail]>
Date: Sat, 22 Jul 2006 13:32:19 +0200
From: Christian Joergensen <[e-mail]>
Organization: razor.dk
MIME-Version: 1.0
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable
To:  [e-mail]
Subject: =?ISO-8859-1?Q???=

Testing. Subject should be: ?

2. Open it using file -> open saved message
Actual Results:  
Subject: =?ISO-8859-1?Q???=

Expected Results:  
Subject: ?

Do not hesitate to contact me if there's any uncertainties in my bug report.

Comment 1

[Christian Joergensen]

Hello? Is there anybody in there? :-)

Comment 2

[Doron Shikmoni]

I'm seeing the same problem. Anytime a question mark appears anywhere within a QP-encoded subject line, the decoding seems to fail and the subject line is presented in its encoded form.

Comment 3

[Dave Miller [:justdave]]

I can reproduce this reliably in Shredder

Comment 4

[Dave Miller [:justdave]]

OS=all because I'm on Mac.

per asuth, the responsible line of code appears to be at
http://mxr.mozilla.org/mailnews/source/netwerk/mime/src/nsMIMEHeaderParamImpl.cpp#691

Comment 5

[Dave Miller [:justdave]]

But I think he really means here:
http://mxr.mozilla.org/comm-central/source/mozilla/netwerk/mime/src/nsMIMEHeaderParamImpl.cpp#660

Comment 6

[Dave Miller [:justdave]]

I suck.

http://mxr.mozilla.org/comm-central/source/mozilla/netwerk/mime/src/nsMIMEHeaderParamImpl.cpp#689

Comment 7

[Dave Miller [:justdave]]

Attachment: Patch v1

Comment 8

[Dave Miller [:justdave]]

Attachment: Patch v2

eh, first one wasn't quite right.

Comment 9

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

Attachment: Patch v3

Justdave was a little too trusting of my previous loop invariant suggestions :)

I will attach a comm-central unit test that exercises this logic next.  It's comm-central because nsIMIMEHeaderParam.idl marks decodeRFC2047Header noscript, and then the comment says that it's internal and MIME_DecodeMimeHeader is the only real consumer (apart from nsIMimeHeaderParam.getParameter).  Since nsIMimeConverter in mailnews/mime exposes this functionality in a scriptable fashion that is straightforward (the method is decodeMimeHeader and explicitly takes a value), it seems like a nice unit testing substrate.  In contrast, nsIMimeHeaderParam.getParameter does all sorts of crazy things.

Comment 10

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

Attachment: comm-central unit test for this

Comment 11

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

biesi, are you the right person to review this for netwerk?

Comment 12

[Jerry Amundson]

biesi, are you there?

Comment 13

[Christian :Biesinger (don't email me, ping me on IRC)]

bugmail is not a good way to reach me, in general. I could review this on Friday, or you can try to find someone else to review this in the meantime.

Comment 14

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

(In reply to comment #13)
> bugmail is not a good way to reach me, in general. I could review this on
> Friday, or you can try to find someone else to review this in the meantime.

We should likely just move the code into comm-central from mozilla-central (cc'ing bienvenu).

On a separate issue, should the module ownership listing for Necko be updated so that someone else is module owner (cc'ing bz)?  (https://wiki.mozilla.org/Modules/Core still lists biesi)

Comment 15

[Boris Zbarsky [:bzbarsky]]

That's up to Christian.  Note that the whole point of having module peers is that the owner may not be able to review everything.  The owner _is_ supposed to have final call on technical issues.

Comment 16

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

Reviewing my comment about the module ownership, I realize it may be ambiguous.  My intent was just to make sure that the module owner listing was accurate, not a statement of opinion about who should or should not be module owner.  While I know some effort was made to clean up the module ownership listings, there are huge swathes of what I understand to be no-man's-lands where I think last-person-who-touched-it or has-any-clue-about-what's-going-on-in-there may have resulted in people being named, perhaps unwillingly.

Having said that, it might helpful to mark those in the listing for the necko module who are more available to perform reviews / are actively checking their review request queue.  I originally issued the review request in 2008 against biesi because I did not know enough about who was active in the project, etc., and assumed that the module owner would be likely to at least bounce the review at someone who could review it.

Comment 17

[Christian :Biesinger (don't email me, ping me on IRC)]

So some comments on module ownership and me :)

There was a period of time where I did not really check my review queue, and that patch fell into that period. Recently I do check my queue and actively do reviews, but there are a fair amount of patches that are still left over in my queue. I really should deal with them in some way.

I feel that I'm still a good choice for module owner, but if people disagree I could change it to someone else.

Comment 18

[Christian :Biesinger (don't email me, ping me on IRC)]

Comment on attachment 334833 [details] [diff] [review]
Patch v3

Review of attachment 334833 [details] [diff] [review]:
-----------------------------------------------------------------

Comment 19

[Patrick McManus [:mcmanus]]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=3ae9bbd6e339

Comment 20

[Patrick McManus [:mcmanus]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/d4fe24eab2b74d044332321e23ba55dab8fdc687
Bug 345580 - Problem decoding quoted-printable question mark in subject r=biesi

Comment 21

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/d4fe24eab2b7