On Tue, 28 Nov 2006 13:58:22 -0800 Dennis Peterson <[EMAIL PROTECTED]> wrote:
> René Bellora wrote: > > > >>>> > >>> This sounded like a good idea, so I gave it a try. After spending a > >>> few hours to script the softlinks I got it to work for small file > >>> lists, but it still doesn't work for lots of files (~5000). When I > >>> run 'clamscan /tmp/clamscan/*' I get the following error: > >>> > >>> /usr/bin/clamscan: Argument list too long > >> > > this could be circumvented with xargs: > > cd /tmp/clamscan > > find . -type f -print0 | xargs -0 clamscan > > Assumes Linux, or at least gnu find and xargs, but also the files are > soft links so the -type f automatically fails. There *are* ways round that -o -type ... > > Also, I believe the OP was interested in finding a way to scan all the > files from a single invocation of clamscan and xargs won't necessarily > do that. No, it's designed to work around the limitations of your OS, and provide the most efficient solution. > > As the requirements have evolved it seems more likely a Perl solution is > most attractive both for creating the list and for logging the results. > And it will eliminate the earlier suggestion of using soft links. This > looks interesting: > http://www.fpsn.net/index.cgi?pg=products&product=File::Scan::ClamAV > > It allows sending files as streams to clamd so there is only a single > invocation of perl and clamd is presumed already running. > > Finally, it is still possible to hack clamscan to read in a file that > contains a list of names of files to scan. > > dp > _______________________________________________ > Help us build a comprehensive ClamAV guide: visit http://wiki.clamav.net > http://lurker.clamav.net/list/clamav-users.html _______________________________________________ Help us build a comprehensive ClamAV guide: visit http://wiki.clamav.net http://lurker.clamav.net/list/clamav-users.html
