Super Search is your friend. read file line in the title for SoPW turns up: Line number in a file, reading a particular line in a large text file, Best way to read line x from a file, Reading Lines from Text file and reading a random line from a file among many many others. This is not an infrequently asked question.

Hi xiaoyafeng,

You have to take a look at Tie::File and perlre.

From the documentation:

A sample code to proceed....

use strict;
use warnings;
use Tie::File;

my @file;

tie @file, 'Tie::File', "test.txt" or die "$!";

for (0..5) {

print "ORA is present\n" if $file[$_] =~ /ORA/;

}
[download]

update: added code.

My question is: How quickly jump to the line I want and begin search?

That's not possible unless you know at which byte offset the desired line starts.

How quickly jump to the line I want and begin search?

Most operating systems do not store text files as records of lines, i.e. they don't store that information. You will need to read the file from the start keeping a count of the number of lines you've read so far. Luckily perl makes this easy for you as long as you read the file line by line (the default):

open my $file,"<","/path/to/file.txt" or die "Can't read file: $!";
while (<$file>) {  # reads one line of text
   # the line string is in $_
   # the line number is in $.
   if ($. <= 10000) {
      print "ORA found on line $.\n" if /ORA/;
   }
   elsif ($. <= 20000) {
      print "Failed found on line $.\n" if /Failed/;
   }
   else {
      print "File found on line $.\n" if /File/;
   }
}
close $file;
[download]