#!/usr/local/bin/perl
#######################################################################
# $Author: jgoerzen $
# $Revision: 1.1.1.1 $
# $Date: 2000/08/19 00:28:56 $
# $Source: /var/cvs/gopher/gopherd/scripts/autoabstract,v $
# $State: Exp $
#
# Paul Lindner, University of Minnesota.
#
# Copyright 1994 by the Regents of the University of Minnesota
# see the file "Copyright" in the distribution for conditions of use.
#######################################################################
# MODULE: autoabstract
# 
# A script to automatically generate a gopher+ ABSTRACT for an item
# It can be set up to build .abstract files too..
# 
#######################################################################
# Revision History:
# $Log: autoabstract,v $
# Revision 1.1.1.1  2000/08/19 00:28:56  jgoerzen
# Import from UMN Gopher 2.3.1 after GPLization
#
# Revision 1.3  1995/02/06  22:27:56  lindner
# Use dynamic space for Data_Dir, remove RunLS
#
# Revision 1.2  1994/12/15  17:34:25  lindner
# Fix for .txt files
#
# Revision 1.1  1994/12/14  19:14:28  lindner
# New scripts for Automatic Gopher Blocks
#
#
#######################################################################


#######################################################################
# Parameters you might want to set..
#
# %txtconvert specifies programs that take non-text output and
#             output a text version.
#
# $cacheit    specifies that this script should automatically generate
#             .abstract files for us.  eats disk space like a pig
#######################################################################
#

$txtconvert{'.ps'}    = "ps2ascii \"%s\"";
$txtconvert{'.html'}  = "lynx -dump \"%s\"";
$txtconvert{'.tar.Z'} = "zcat \"%s\" | tar tvf -";
$txtconvert{'.zip'}   = "unzip -l \"%s\"";
$txtconvert{'.tex'}   = "detex %s";

$cacheit = 0;

#
#######################################################################


$filenamebase = shift || die "Need a filename...";
$fname = $filenamebase;

$ENV{'PATH'} = $ENV{'PATH'} . ':/usr/local/bin:/usr/gnu/bin';
#
# First, find out what filenames there are..
#

exit if (-f "$fname.abstract");
exit if (-d "$fname");
$fname = &FindFiletoAbstract($fname);

($dev,$ino,$mode,$nlink,$uid,$gid,$rdev,$size, $atime,
 $mtime,$ctime,$blksize,$blocks) = stat($fname);
exit if ($size < 512);


exit if ($fname =~ /gif$/i);

open(f, "<$fname");

foreach $e (keys(%txtconvert)) {
    if ($fname =~ /$e$/i) {
	$cmd = sprintf($txtconvert{"$e"}, $fname);
	close(f);
	open(f, "$cmd |");
	done;
    }
}




#
# Slurp file a paragraph at a time..
#
$/ = "";

#
# Read the first paragraph, lots of interesting stuff here..
#

$_ = <f>;
while (! (/^[\s]*/)) {
    $_ = <f>;
}

$firstpara = $_;

if (/^From [\S]+[\s]+\S\S\S[\s]+\S\S\S[\s]+[\d]+ [\d]+:[\d]+:[\d]+ \d\d\d\d/ ||
    /^Newsgroups: / || /^Article[\s]+[\d]+[\s]+of/ || /^Path: /) {
    # News or mail thingie...
    
    @lines = split(/\n/);

    foreach $i (@lines) {
	if ($i =~ /^Subject: (.*)$/) {
	    $abstract .= $1 . "\n";
	}
	if ($i =~ /^Newsgroups: (.*)$/) {
	    $abstract .= $i . "\n";
	}
	if ($i =~ /^Organization: (.*)$/) {
	    $abstract .= $i . "\n";
	}
	if ($i =~ /^Summary: (.*)$/) {
	    $abstract .= $1 . "\n";
	}

	if ($i =~ /^From: (.*)$/) {
	    $abstract .= $i . "\n";
	}
	
	if ($i =~ /^Keywords: (.*)$/) {
	    $abstract .= $1 . "\n";
	}
    }

}
elsif (/^\#!\/.*perl/ || /^\#!\/.*sh/) {
    @lines = split(/\n/);

    shift(@lines);
    foreach $i (@lines) {
	if (/\#/) {
	    $abstract .= "$i\n";
	}
    }
}

else {

    while (<f>) {
	if (length($firstpara) < 300) {
	    $firstpara .= "\n$_";
	}
	if (/^[\s]+Abstract/i) {
	    $abstract .= $_;
	}
	if ($lastpara =~ /Abstract[\s]+$/i) {
	    $abstract .= $_;
	}

	if (/^[\s]+Summary/i) {
	    $abstract .= $_;
	}
	if ($lastpara =~ /Summary[\s]+$/i) {
	    $abstract .= $_;
	}

	$lastpara = $_;
    }
}
$len = length($abstract);


exit if ($len == $size);

$abstract = $firstpara if ($len == 0);

print "$abstract\n";

if ($cacheit) {
    open(abstractout, ">$filenamebase.abstract");
    print abstractout "$abstract\n";
    close(abstractout);
}


sub FindFiletoAbstract {
    local($fname) = @_;
    local($base, $f);

    return($fname) if ( -f $fname);

    #
    # Probably not a text file...
    #

    $fname =~ m/^(.*)\/([^\/]+)$/;
    $base = $1;
    $f = $2;

    
    if (!($fname =~ /\//)) { 
	$base = ".";
	$f = $fname;
    }


    opendir(thedir, $base);
    @files = readdir(thedir);
    foreach $i (@files) {
	if ($i =~ /^$f/) {
	    foreach $t (keys(%txtconvert)) {
		return("$base/$i") if ($i =~ /$t$/i);
	    }
	    return("$base/$i") if ($i =~ /txt$/i);
	    push(@ourfiles, $i) 
	}
    }
    closedir(thedir);

    return($ourfiles[1]);
}