From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jan Kara Subject: Re: [PATCH 6/8] Support non-BMP characters in UDF Date: Wed, 16 May 2012 22:04:59 +0200 Message-ID: <20120516200459.GD1687@quack.suse.cz> References: <4FB2E25E.900@gmail.com> <20120516143448.GD27661@quack.suse.cz> <4FB3C44F.6080409@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Jan Kara , linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org To: Vladimir =?utf-8?Q?'=CF=86-coder=2Fphcoder'?= Serbinenko Return-path: Received: from cantor2.suse.de ([195.135.220.15]:39601 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932429Ab2EPUFM (ORCPT ); Wed, 16 May 2012 16:05:12 -0400 Content-Disposition: inline In-Reply-To: <4FB3C44F.6080409@gmail.com> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On Wed 16-05-12 17:14:23, Vladimir '=CF=86-coder/phcoder' Serbinenko wr= ote: > On 16.05.2012 16:34, Jan Kara wrote: > > On Wed 16-05-12 01:10:22, Vladimir '=CF=86-coder/phcoder' Serbinenk= o wrote: > >> I also have a counterpart for mkudffs/udf-tools but sourceforge ho= mepage > >> seems to be abandoned does anybody know if there is a new homepage= for > >> mkudffs? Oh, and I forgot to reply here: mkudffs is really unmaintained. But a= lso it's not used too much AFAIK. Most people use genisoimage to generate u= df filesystems. > > Thanks for the patch! >=20 > You're welcome. Thanks for reviewing. >=20 > It looks OK but shouldn't we rather use the helper > > functions you introduced in the NLS code? It look wrong to replicat= e > > decoding of UTF16 here. > >=20 >=20 > The helper functions are limited to buffers aligned on 16-bit boundar= y > which is not the case of this buffer. I see following solutions: I see. > 0) Homegrown like in previous patch > 1) Add a new "endianness" UTF16_LITTLE_ENDIAN_UNALIGNED > 2) Split code for "compressed" vs "uncompressed" and copy the string = to > a temporary buffer in "uncompressed" branch. > 3) Like 2 but make buffer sliding and contain only 2 elements. >=20 > I think 1 or 3 would be the most reasonable. Which solution do you pr= efer? I think 1 would be the best since then it can be easily reused by oth= er filesystems which may have similar issue. Honza --=20 Jan Kara SUSE Labs, CR -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel= " in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html