Path: utzoo!attcan!uunet!munnari.oz.au!uhccux!ames!mailrus!tut.cis.ohio-state.edu!purdue!haven!uvaarpa!mmdf
From: martin%EASBY.DURHAM.AC.UK@CORNELLC.cit.cornell.edu (Martin Ward)
Newsgroups: comp.lang.perl
Subject: (none)
Message-ID: <1990Jun1.205310.23678@uvaarpa.Virginia.EDU>
Date: 1 Jun 90 20:53:10 GMT
Sender: mmdf@uvaarpa.Virginia.EDU (Uvaarpa Mail System)
Reply-To: martin%EASBY.DURHAM.AC.UK@CORNELLC.cit.cornell.edu
Organization: The Internet
Lines: 22


I have a small problem which I would like some perl wizards' help on (or
perhaps it has already been solved). I would like to write a file "viewer",
based on "less", which can cope with more than just ascii files in a sensible
way. For example a compressed file should be piped through "uncompress"
before viewing, a QED file should be passed through "qtoa", nroff files
passed though nroff (with the appropriate macro package guessed at from
the file?) etc.  (other possibilities: use "tar -tf" on a tar file, "strings"
on an executable and so on). A major requirement is that it should produce
the first page of output quickly (ie using pipes rather than uncompressing
etc. the whole file). So for example a compressed nroff file in QED editor
form will be automatically passed to "uncompress | qtoa | nroff -man | less -s"
Most of the file types can be gleaned from the magic no. or the first few lines
of the file, which can then be re-read from the beginning and passed to the
"pipe so far". But what if the input is from a pipe? It looks like we need
some kind of "unread" or "putback" function. Ideas anyone?

			Martin.

My ARPANET address is:  martin%EASBY.DUR.AC.UK@CUNYVM.CUNY.EDU
OR: martin%uk.ac.dur.easby@nfsnet-relay.ac.uk  UUCP:...!mcvax!ukc!easby!martin
JANET: martin@uk.ac.dur.easby    BITNET: IN%"MARTIN@EASBY.DURHAM.AC.UK"