Grabbing Chunks of Data from File in Perl

Question

I am scanning a error log for the keyword "warning." When I find it, there is a particular case of warning that will end information relevant to the warning with only a newline character. What I want to do is scan through this error log, find the keyword warning keep skipping down the file until I find ONLY a newline character and then once I've found it, analyze that chunk and print it out.

The format would be something like this:

Warning! blah
blah
blah
orange

Warning! blah
blah
red
blah

Warning!
blah
blah
blah
yellow

Here's my code so far: use strict; use warnings;

my $file = $ARGV[0];
my $count = 0;

open(LOG, "<",$file) or die "cannot open < input: $!"; 

my $line = <LOG>;
while($line ne "")
{
if($line =~ /warning/i)
{
    while($line !~/\A\n\z/)
    {
        if($line =~ /red/ || $line =~ /yellow/ ||$line =~ /black/)
        {
            $count++; 
        }
        print($line);
        $line = <LOG>; 
    }
    print("Just finished a chunk.\n");
}

$line = <LOG>;
}
print($fix, "\n");

However, it's an infinite loop that constantly throws errors and I don't know why.

Answer 1

You could read file by "chunks" and not by lines

local $/ = "\n\n";
while(my $chunk = <LOG>) { 
  .. 
  print("Just finished a chunk.\n");
}

Answer 2

You enter an infinite loop because when you reach the end of file and $line is undefined, the expression

$line !~/\A\n\z/

is still true. You could try something like

    ...
    while (defined($line) && $line !~ /\A\n\z/) {
       ...

instead.