Re: [GIT PULL] PCI fixes for v4.15

[Peter Grayson]

On Wed, Jan 24, 2018 at 08:40:50AM -0800, Linus Torvalds wrote:
> On Wed, Jan 24, 2018 at 8:20 AM, Bjorn Helgaas <helgaas@xxxxxxxxxx> wrote:
> >> Bjorn, maybe you can send Catalin an example mbox?
> >
> > Attaching the one I used above.
> 
> Heh. That's a mess. It has
> 
>     Content-Type: text/plain; charset=UTF-8
> 
> but then the name in the body is actually Latin1-encoded if I read it correctly.
> 
> Git will auto-convert invalid utf-8 by assuming it is latin1, so it
> all ends up working, but Christian did something wrong in his mailer
> too.

The latest stgit release (v0.18) ignores any mis-encoding of the email
body. However, stgit master now decodes email bodies and is thus exposed
to this kind of stray latin-1 character in a UTF-8 body.

I believe stgit's goal should be to identify and repair this kind of
issue as git does. I will be working on that.

> I suspect that the reason stgit screws up is the quoting of the name:
> 
>    From: "=?UTF-8?q?Christian=20K=C3=B6nig?=" <ckoenig.leichtzumerken@xxxxxxxxx>
> 
> maybe stgit thinks that quoting means "no charset translation".

Thank you for pointing-out this quoting. You are correct that it is part
of the problem for stgit.

The Python2 email module seems to treat the encoded words within the
quotes literally, which leads to the problem at-hand where the encoded
words are being mapped to the Author field of the git commit.

> But I'll leave it to Catalin & co to figure out. The last stgit commit
> was just a few days ago, so it seems to be maintained.
> 
> Bjorn: it might be worth it trying to the very latest stgit:
> 
>     git://repo.or.cz/stgit.git
> 
> because there are _some_ locale changes in there.

Unfortunately, the head of stgit master does not yet solve this issue. I
am working to remedy that.

Pete