Explain this perl code which displays common lines in 2 files

Question

How does this perl one-liner display lines that 2 files have in common?

perl -ne \'print if ($seen{$_} .= @ARGV) =~ /10$/\'  file1 file2

Answer 1

@ARGV is shifted when opening the first file. In scalar context, it now returns 1 (because it has one member, the second file). For each line, this 1 is appended to the hash %seen. When the second file is opened, @ARGV is shifted again and is now empty, so returns 0 in the scalar context. /10$/ means "the line was seen in file1 and now it has been seen in file2 for the first time".

Answer 2

The -n command line option transforms the code to something equivalent to

while ($ARGV = shift @ARGV) {
  open ARGV, $ARGV;
  LINE: while (defined($_ = <ARGV>)) {
    $seen{$_} .= @ARGV;
    print $_ if $seen{$_} =~ /10$/;
  }
}

While the first file is being read, scalar @ARGV is 1. For each line, 1 will be appended to the %seen entry.

While the second file is being read, scalar @ARGV is 0. So if a line was in file 1 and in file2, the entry will look like 1110000 (it was 3× in file1, 4× in file2).

We only want to output common lines exactly one time. We do this when a common line was first seen in file2, so $seen{$_} is 1110. This is expressed as the regex /10$/: The string 10 must appear at the end.