Something

sub get_text {
        my @a = split /\./,"$_[0]";
        $type = $filetypes{"$a[1]"};
        unless ($type) {
                $type = 'null';
        }
        #my $target = "$_[1].$type";
        #scp from remote $_[0] to $target
        my $target=$_[0];
        my $return = '';
        if ($type eq 'pdf') {
                #system("pstotext $target $tempfile");
                exec("/usr/local/bin/pdftotext $target $tempfile");
                open(INFILE, "$tempfile");
                @text=<INFILE>;
                close(INFILE);
                $return=join("",@text);
#               $return='';
        }
        elsif ($type eq 'html') {
                $return=`html2text $target`;
        }
        elsif ($type eq 'xls') {
                $return=`xls2csv $target`;
        }
        elsif ($type eq 'doc') {
                $return=`catdoc $target`;
        }
        elsif ($type eq 'txt') {
                open(INFILE, "$target`");
                @text=<INFILE>;
                close(INFILE);
                $return=join("",@text);
        }
        else {
                $return = '';
        }
        #$content=~s/[^!-~]/ /g;
        $return=~s/[^$list]/ /g;  #replace all control characters with a space.
        $return=~s/\'/\'\'/g;  #escape all single quotes since they'll mess up the SQL.
        $return=~s/\s+/ /g;  #replace all instance of multiple spaces with single spaces.
        @contents=();
        $flag=0;
        $m=0;
        $nextstring=$return;
        $lastspace=0;
        while (!$flag) {
                $lastspace = rindex($nextstring,' ',2000);  #this returns the last instance of a space before 2000 chars
                $contents[$m]=substr($nextstring,0,$lastspace);  #grabs the string until the last instance of a space before 2000 chars
                $nextstring=substr($nextstring,$lastspace+1);  #grabs the string from the last space to the end.
                if (length($nextstring)<2000) {
                        $contents[$m+1]=$nextstring;
                        $flag=1;
                }
                $m++;
        }

}