Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!swrinde!cs.utexas.edu!chinacat!woody
From: woody@chinacat.unicom.com (Woody Baker @ Eagle Signal)
Newsgroups: comp.lang.postscript
Subject: Re: "Take out" a postion of a PS file?
Keywords: Postscript
Message-ID: <1991Jun2.051004.11517@chinacat.unicom.com>
Date: 2 Jun 91 05:10:04 GMT
References: <.675678144@dutepp1>
Organization: a guest of Unicom Systems Development, Austin
Lines: 36

In article <.675678144@dutepp1> luo@dutepp1.et.tudelft.nl (Jian Luo) writes:
>
>Somebody has developed such a utility?  For example, this should work like:
>ps2ps -f5 -t10 input.file, output.file
>which will put the pages 5 to 10 to the output file (Postscript too).

As has been pointed out numerous times, this is more than just a trivial
problem.  Since keywords can be redefined, and aliases created for them,
it becomes almost impossible to determine where a page ends.  The best solution
to this problem, is an analysis of the output file, looking through the prolog
for the end of page command that it uses.  (create a couple of single line
pages) and then once you have found that, find the start of the page code
the same way.  Then write a routine to scan the file, locating the start of
page and end of page to isolate each page.  be sure and emit the preamble.
I recently did this for a company here in town for the INTERLEAF package
running on a sun workstation.  They needed to take the output from interleaf
and create a file stream that had the pages rearranged such that they
would print 2 up on both sides of a sheet (using a KODAK high speed
PS printer) such that when the pages were collated, and staped they would
create a neat booklet.  It took about 16 hours of hacking and work to get it
perfect, but it surely works fine for them.  The code is however, specific
to the interleaf software.

That is the surest way to accomplish this.  A good shell script programmer, 
using standard unix tools such as grep and awk should be able to create
a shell script to do this, again for a specific package.  Personally,
I have been toying with the idea of making a general routine that would
read a data file to retrieve lines of data, and use it to parse out
pages.  the file would contain the start of preamble line, the end of 
preamble line, the start of page line and the end of page line.  That way
one could just create this file, and bingo, you have one of these for
any given application package.

Cheers
Woody
p.s.  I HAVE NOT written this, and am a tad to busy to do so.