From: Marco Costalba <mcostalba@gmail.com>
Date: Tue, 02 Jan 2007 22:29:14 GMT
Subject: Re: Possible regression in git-rev-list --header
Message-ID: <e5bfff550701021429j1587949fgfe19effcaf74e6df@mail.gmail.com>
In-Reply-To: <7vps9xrs9y.fsf@assigned-by-dhcp.cox.net>

On 1/2/07, Junio C Hamano <junkio@cox.net> wrote:
>
> > And Marco has to fix the header parsing anyway.
>
> No question about that.  If iconv() punts, qgit will see
> "encoding" header to deal with even when the re-coding is in
> effect.  I think it may be a sensible thing for qgit to replace
> the log message and show "log message in this encoding, which
> cannot be shown in this window" instead in such a case, but that
> is up to Porcelain.
>

Yes, git-rev-list output is stored in memory as a big chunk of row
bytes, conversion
to utf8 (the internal format used by Qt string class) is done _only_
for displayed items, just
before to show them to user so is not performance critical. I agree
with Junio, that info could be used to teach Qt string conversion
function how to handle the data. Something like:

         convertWithCodec(const char* data, codec* codecName)

So perhaps if Qt is better then iconv we could display a good log
message anyway.