Although hbm makes a good point, in case you do want to delete duplicate xml lines, here's one way:

use Modern::Perl;

my %seen;
say for grep { chomp; !$seen{$_}++ } <DATA>;

__DATA__
<item pr_1="a" pr_2="2" internal="co_1" pr_family="2" />
<item pr_1="b" pr_2="2" internal="co_1" pr_family="2" />
<item pr_1="c" pr_2="2" internal="co_1" pr_family="2" />
<item pr_1="a" pr_2="1" internal="co_1" pr_family="2" />
<item pr_1="b" pr_2="1" internal="co_1" pr_family="2" />
<item pr_1="c" pr_2="1" internal="co_1" pr_family="2" />
<item pr_1="a" pr_2="2" internal="co_1" pr_family="2" />
<item pr_1="b" pr_2="2" internal="co_1" pr_family="2" />
<item pr_1="c" pr_2="2" internal="co_1" pr_family="2" />
[download]

Output:

<item pr_1="a" pr_2="2" internal="co_1" pr_family="2" />
<item pr_1="b" pr_2="2" internal="co_1" pr_family="2" />
<item pr_1="c" pr_2="2" internal="co_1" pr_family="2" />
<item pr_1="a" pr_2="1" internal="co_1" pr_family="2" />
<item pr_1="b" pr_2="1" internal="co_1" pr_family="2" />
<item pr_1="c" pr_2="1" internal="co_1" pr_family="2" />
[download]

Hope this helps!