Ignore lines with same next fields as previous

Question

I have a file contents of which looks like this

123,1,ABC,DEF
123,1,ABC
345,4,TZY
456,3,XYZ
333,4,TTT,YYY
333,4,TTT

I want to ignore lines with the previous and next contents the same i.e lines containing 123 and 333

Output needs to be

345,4,TZY
456,3,XYZ

Any ideas on how to go about this

Answer 1

try the uniq utility

uniq -w 3 your_file.txt

would do the trick. no need for perl

Answer 2

TMTOWTDI:

my $str = join '', <>;
$str =~ s/^(\d+).+\n(\1.+\n)+//mg;
print $str;

EDIT: first line can also be replaced with Randal L. Schwartz's slurp:

my $str = do { local $/; <HANDLE> }; #

Answer 3

TMTOWDI

my $last_prefix = ""; 
my $last_line = ""; 
while (<>) { check_line($_); }
check_line("");    sub check_line {
    my $line = shift;
    my ($prefix) = ($line =~ /^([^,]*),/); 
    if (($prefix || "") ne $last_prefix ) {
        print $last_line;
        $last_line = $_;
    } else {
        $last_line = "";
    };
    $last_prefix = $prefix; 

}

This is wordy but I suspect the performance might be better than regexp on a very large file.