For the first question: you are dealing with a multiline match, in the sense that after name is found any num belong to it. This is achieved better with something like $current_name initialized out of the loop. you were right with the `push @{hash1{name1}},$x1,$y1,$x2,$y2` tecnique but you missed a dollar sigyl before name1: in your actual code with the statement $hash1{$name1}=[$x1,$y1,$x2,$y2]; you are reinizialing the value to a new array each time.

Notice that I used push @{$hash1{$current_name}}, $_ for $line =~ /\d+/g pushing all numbers into the array: perhaps you want more robust control appending a check like: if $line =~/^num\s+\d/

#!/usr/bin/perl
use strict;
use warnings;
use Data::Dumper;

my $current_name;
my %hash1;
while (my $line = <DATA>){
    chomp $line;
    if($line =~ /^name\s+(\S+)/)    {$current_name = $1;}
    push @{$hash1{$current_name}}, $_ for $line =~ /\d+/g
}
print Dumper \%hash1;
__DATA__
name    foo
num     111 222 333 444
name    jack
num     999 111 222 333
num     333 444 555 777
[download]

See the above running at webperl demo by haukex (oh oh oh!;)

L*

UPDATE

for the second question you can profit of the exists function: foreach key of the first hash, if exist the same in the second one, you print also the second list. If you want values in group of four it's a bit trickier but feasible

UPDATE October 22 in reponse to Re^2: how to push multiples row of values into hash and do comparison

$_ is the default input variable in perl: see perlvar

In my code I use it only in the push @{$hash1{$current_name}}, $_ for $line =~ /\d+/g statement. Let's reduce the example a bit. The g modifier in the regex in list context returns the list of all matches: see it in perlrequick so $line =~ /\d+/g returns a list.

The first part push @{$hash1{$current_name}}, $_ push into the array the default variable of the foreach loop aka $_ so translated to english sounds like: push in this array every match obtained matching one or more number in this line.