This is one of the easiest tasks for Perl. Contrary to the one-liner that would do, some more verbose code:

# let's define a sub that will return the matched id IF there was a ma
+tch, undef else
sub match_id {
   my $str = shift;

   if ($str =~ m{dp/(.\d+?)\z}) {  # we assume our id is "some charact
+er followed by digits"
       return "$1\n";  # will return id IF we had a match
   }

   return;  # undefined if there was no match
}

# assume you have these strings one in a line - or not
# and read the file to a list (one line = one list element)
# then you can get a list of matches like so:

my @ids = grep { defined $_ } map { match_id($_) } @lines_from_file;
[download]

If you have the strings scattered around your file, possibly several on one line, you better use the global modifier and a while loop. You also should tighten your regex then - specifying the ID more exact:

while ($file_content =~ m{dp/(.\d+?)}g) {  # we have to omit the \z
  print "Found one: $1\n";
}
[download]

E.g. if you knew, all your Id's started with an upper-case letter and have always 8 digits, that would be

m{dp/([A-Z]\d{8})}g
[download]