Re: Balsa port to gmime-2.4 regression

[Jeffrey Stedfast <fejj novell com>]

Pawel Salek wrote:
> On 01/01/2009 11:36:33 AM, Ildar Mulyukov wrote:
>> See attached message. It's subject shown as
>> [Bug
>> 18260]=?UTF-8?Q?=20=D0=BD=D0=B5=D1=82=20=D0=B7=D0=B0=D0=B2=D0=B8=D1=81=D0=B8=D0=BC=D0=BE=D1=81=D1=82=D0=B8=20=D0=BD=D0=B0=20libstrongswan?=
>>
>>
>> Balsa with gmime-2.0 and mutt shows allright.
>> Codebase:
>> http://git.altlinux.org/people/ildar/packages/?p=balsa.git;a=shortlog;h=refs/heads/patches/gmime-2.4
>>
>>
>> Anyone to help?
>
> So, the question is whether the following string is a properly encoded
> or not:
>
> 0]=?UTF-8?Q?=20=D0=BD=D0=B5=20libstrongswan?=
>
> According to rfc2822, value of Subject header is supposed to be an
> unstructured text but I do not see directly how this connects to
> encoded words of RFC2047... Anybody?
>
> Pawel
>
Technically, that encoding is broken but I thought I had added logic to
GMime to try and work around issues like that. In GMime 2.2.x, this
workaround was enabled via a compile-time configure switch, but in 2.4.x
it is enabled via a runtime flag to g_mime_init() called
GMIME_ENABLE_RFC2047_WORKAROUNDS.

I changed this for 2 reasons:

1. so that different software can decide on using a different option
using the same installed version of gmime

2. (and more importantly) it makes it easier for me to unit test both
cases :)

Jeff