DMARC-Filter: OpenDMARC Filter v1.3.2 sourceware.org BD418384C00B
Subject: Re: "ls" sorts wrongly if given large number of files
To: cygwin@cygwin.com
References: <DB8P191MB06363D09B9C9E54A289AAB7AB4D00@DB8P191MB0636.EURP191.PROD.OUTLOOK.COM>
Message-ID: <fc02f42c-ae7d-997f-b652-de9ffe243329@gmx.net>
Date: Wed, 6 Jan 2021 19:41:56 +0100
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:78.0) Gecko/20100101
 Thunderbird/78.6.0
MIME-Version: 1.0
In-Reply-To: <DB8P191MB06363D09B9C9E54A289AAB7AB4D00@DB8P191MB0636.EURP191.PROD.OUTLOOK.COM>
Content-Language: de-DE
Precedence: list
From: Frank-Ulrich Sommer via Cygwin <cygwin@cygwin.com>
Reply-To: Frank-Ulrich Sommer <f-u.s@gmx.net>
Content-Type: text/plain; charset="utf-8"; Format="flowed"
Errors-To: cygwin-bounces@cygwin.com
Sender: "Cygwin" <cygwin-bounces@cygwin.com>
Content-Transfer-Encoding: 8bit


Am 06.01.2021 um 19:17 schrieb Kamran via Cygwin:
> Hi all
>
> "ls" (version 8.26) sorts wrongly if given large number of files via "find" or "xargs"
>
> For example:
>
> find -type f -exec ls -oS -h {} +
>
> OR
>
> find -type f -print0 | xargs -0r ls -oS -h
>
> Gives following data. Sorry for the long listing, and wrapped lines. But search for "setup.ini" which is about 17 MB, it is sorted AFTER very small files.
>
> In fact it seams that sorting is restarted from that file (subsequent files are again sorted). Note also that removing "-h" from "ls" command lines results in the same problem.
>
> (output is trimmed to remove unnecessary data, i.e. perms/user/date-time, but order is the same)
>
>  26M ./release/gcc/gcc-core/gcc-core-10.2.0-1.tar.xz
>  24M ./release/binutils/binutils-2.34+1git.de9c1b7cfe-1.tar.xz
> [...]
> 108 ./release/python2/python/python-2.7.16-1.tar.xz
>  108 ./release/python-gobject/python-gobject-2.28.7-1.tar.xz
>   46 ./release/tcl-tk/tcltk/tcltk-20120206-1.tar.bz2
>   32 ./release/man-db/man/man-2.6.7-2.tar.xz
>   32 ./release/popt/popt-1.16-2.tar.xz
>   32 ./release/procps-ng/procps/procps-3.3.10-1.tar.xz
>  17M ./setup.ini
> 5.7M ./release/vim/vim-common/vim-common-8.2.0486-1.tar.xz
> 4.6M ./release/w32api-headers/w32api-headers-8.0.0-1.tar.xz
> 1.8M ./release/vim/vim-doc/vim-doc-8.2.0486-1.tar.xz
> [...]
> 16K ./release/xeyes/xeyes-1.1.2-1.tar.xz
>  15K ./release/xf86-video-dummy/xf86-video-dummy-0.3.8-1.tar.xz
>  12K ./release/util-linux/libuuid1/libuuid1-2.33.1-2.tar.xz
>
xargs uses multiple calls to ls and find calls ls once for each matching file so in the two mentioned cases ls will not get to see the full list and thus can't sort all the files. This is the expected bahviour and not special to cygwin.
--
Problem reports:      https://cygwin.com/problems.html
FAQ:                  https://cygwin.com/faq/
Documentation:        https://cygwin.com/docs.html
Unsubscribe info:     https://cygwin.com/ml/#unsubscribe-simple