@seen{@A} = ();
is the same as

$seen{$A[0]}) = undef;
$seen{$A[1]}) = undef;
$seen{$A[2]}) = undef;
$seen{$A[3]}) = undef;
...
[download]

@merged = (@A, grep{!exists $seen{$_}} @B);
is the same as

@C = grep{!exists $seen{$_}} @B;
@merged = (@A, @C);
[download]

@C = grep{!exists $seen{$_}} @B;
@C is assigned the contents of @B not seen in %seen. In other words,
@C is assigned the contents of @B not seen in @A.

Hey and what would $seen{$_} do ???
$_ = each value of A and then compare it with each B ?
AM i right?
Thanks
Krishna

$_ takes each value of **@B**, since it's grep {...} @B

@A = (1, 4, 7, 10, 14);
@B = (1, 3, 2, 4);

@seen{@A} = ();
# does:
# $seen{ 1} = undef;
# $seen{ 4} = undef;
# $seen{ 7} = undef;
# $seen{10} = undef;
# $seen{14} = undef;

@C = grep { !exists($seen{$_}) } @B;
# In the grep evaluator, $_ takes the value of each
# element of argument list, one at a time.
#
# !exists($seen{1}) evaluates to false (undef).
# !exists($seen{3}) evaluates to true  (1).
# !exists($seen{2}) evaluates to true  (1).
# !exists($seen{4}) evaluates to false (undef).
#
# grep returns the values that resulted in true,
# so the result is:
# @C = (3, 2);

@merged = (@A, @C);
# @merged = (1, 4, 7, 10, 14, 3, 2);
#          @A---------------  ----@C
[download]