Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!bloom-beacon!mit-eddie!killer!vaxnix!merch!cpe!tif
From: tif@cpe.UUCP
Newsgroups: comp.sources.wanted
Subject: Re: Unique set of file names/inodes?
Message-ID: <131400004@cpe>
Date: 22 Feb 89 14:17:00 GMT
References: <822@n8emr.UUCP>
Lines: 17
Nf-ID: #R:n8emr.UUCP:822:cpe:131400004:000:670
Nf-From: cpe.UUCP!tif    Feb 22 08:17:00 1989


Written  2:10 pm  Feb 20, 1989 by n8emr.UUCP!lwv in cpe:comp.sources.w
>I want to take the output of a find dir -criteria_of_various_sort -print and
>massage it into a smaller list of file names which have no overlap.  Files

Off the top of my head, try something like this

find dir -whatever -print | xargs ls -i | sort -n -u -2 | sed 's/^[^ ]*//'

At least this eliminates the System V style links (I don't know anything
about other types of links) by sorting uniquely on the inode number.

Note:  It won't work quite right if find goes across to a different filesystem.

			Paul Chamberlain
			Computer Product Engineering, Tandy Corp.
			{killer | texbell}!cpe!tif