Zammad: UTF-8 encoded Email subjects like "=?UTF-8?Q? Personal=C3=A4nderung?=" not decoded

Created on 29 Jan 2019  ·  3Comments  ·  Source: zammad/zammad

Infos:

  • Used Zammad version: 2.8
  • Installation method (source, package, ..): any
  • Operating system: any
  • Database + version: any
  • Elasticsearch version: any
  • Browser + version: any

Expected behavior:

  • the parsed subject should be Personaländerung

Actual behavior:

  • the parsed subject is =?UTF-8?Q?Personal=C3=A4nderung?=

Steps to reproduce the behavior:

  • Parse an Email with a subject header that's an UTF-8 quoted printable string like =?UTF-8?Q?Personal=C3=A4nderung?=

Yes I'm sure this is a bug and no feature request or a general question.

This is related to #1045, which is about parsing a us-ascii quoted printable string subject.

EDIT: The original report contained an incorrect triggering string: =?UTF-8?Q?Personal=C3=A4nderung?=. Thanks to @martinvonwittich for pointing out the mistake. The correct triggering string is =?UTF-8?Q? Personal=C3=A4nderung?=, with an added significant space character.

bug mail processing verified

Most helpful comment

Thanks for double checking, @martinvonwittich! You are correct, I made a mistake in the bug report. The offending string is =?UTF-8?Q? Personal=C3=A4nderung?=, with a space in front of the string. The original subject was a sentence containing the word "Personaländerung", with a space in front of the word itself. I accidentally stripped the space in front off, thinking that it's purely the umlaut causing the problem. But alas, the extra space is actually critical in activating the bug.

I've edited the issue title and my OP, along with an edit note.

All 3 comments

For some reason I'm unable to reproduce this. My Zammad 2.8.x instance seems to handle both these cases fine:

  • Subject: =?utf-8?Q?Personal=C3=A4nderung?=
  • Subject: =?UTF-8?Q?Personal=C3=A4nderung?=

The complete mail source:

Return-Path: <redacted@redacted>
Received: from redacted (localhost [127.0.0.1])
     by redacted (Cyrus 2.5.10-Debian-2.5.10-3+1~iserv10809+10869-1) with LMTPA;
     Wed, 30 Jan 2019 00:02:32 +0100
X-Sieve: CMU Sieve 2.4
Received: from localhost ([127.0.0.1] helo=redacted)
    by redacted with esmtpa (Exim 4.89)
    (envelope-from <redacted@redacted>)
    id 1gocOm-0007PR-Cy
    for redacted@redacted; Wed, 30 Jan 2019 00:02:32 +0100
Message-ID: <9a21f01195853f35cb86c66c64bb2d5c@redacted>
Date: Wed, 30 Jan 2019 00:02:32 +0100
Subject: =?UTF-8?Q?Personal=C3=A4nderung?=
From: Martin von Wittich <redacted@redacted>
To: zammad <redacted@redacted>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
X-Mailer: IServ v3

test

Thanks for double checking, @martinvonwittich! You are correct, I made a mistake in the bug report. The offending string is =?UTF-8?Q? Personal=C3=A4nderung?=, with a space in front of the string. The original subject was a sentence containing the word "Personaländerung", with a space in front of the word itself. I accidentally stripped the space in front off, thinking that it's purely the umlaut causing the problem. But alas, the extra space is actually critical in activating the bug.

I've edited the issue title and my OP, along with an edit note.

The previously incorrect test cases have been fixed via 0b472fa.

Was this page helpful?
0 / 5 - 0 ratings