Taking only values which form continous range - perl

Question

I have a file with 3 columns ->
A1 0 9
A1 4 14
A1 16 24
A1 25 54
A1 64 84
A1 74 84
A2 15 20
A2 19 50
I want to check if each line (value in col2 and 3) is present already or is in between the range of previous line, if col1 value is equal.
The desired output is ->
A1 0 14
A1 16 54
A1 64 84
A2 15 50
I have tried ->
#ARGV or die "No input file specified";
open $first, '<',$ARGV[0] or die "Unable to open input file: $!";
#open $second,'<', $ARGV[1] or die "Unable to open input file: $!";
$k=0;
while (<$first>)
{
if($k==0)
{
#cols = split /\s+/;
$p0=$cols[0];
$p1=$cols[1];
$p2=$cols[2];
$p3=$cols[2]+1;
}
else{
#new = split /\s+/;
if ($new[0] eq $p0){
if ($new[1]>$p3)
{
print join("\t", #new),"\n";
$p0=$new[0];
$p1=$new[1];
$p2=$new[2];
$p3=$new[2]+1;
}
elsif ($new[2]>=$p2)
{
print $p0,"\t",$p1,"\t",$new[2],"\n";
$p2=$new[2];
$p3=$new[2]+1;
}
else
{
$p5=1;
}
}
else
{
print join("\t", #new),"\n";
$p0=$new[0];
$p1=$new[1];
$p2=$new[2];
$p3=$new[2]+1;
}}
$k=1;
}
and output I am getting is ->
A1 0 14
A1 16 24
A1 16 54
A1 64 84
A1 64 84
A2 15 20
A2 22 50
I am not able to understand why I am getting this wrong output. Also if there is any way that I can erase(or overwrite) the last printed line, then it will be very easy.

Answer 1

First of all, it would be much more simple to help you if you
used strict and warnings, and declared all your variabled close to first use with my
indented your code properly to show the structure
The reason your code fails is that you are printing data under too many conditions. For example you output A1 16 24 when you find it cannot be joined with the previous range A1 4 14 without waiting for it to be extended by the subsequent A1 25 54 (when you correctly extend the range and print it again). A1 64 84 is output twice for the same reason: first because it cannot be merged with A1 25 54, and again because it has been "extended" with A1 74 84. Finally A2 15 20 is output straight away because it has a new first column, even though it is merged with the next line and output again.
You need to output a range only when you have found that it cannot be extended again. That happens when
a new record is found that doesn't overlap the existing data
the end of the file is reached
This code prints output only in those cases an appears to do what you need.
use strict;
use warnings;
my #data;
while (<DATA>) {
if (not #data) {
#data = split;
next;
}
my #new = split;
if ($new[0] eq $data[0] and $new[1] <= $data[2] + 1) {
$data[2] = $new[2];
}
else {
print join("\t", #data), "\n";
#data = #new;
}
print join("\t", #data), "\n" if eof DATA;
}
__DATA__
A1 0 9
A1 4 14
A1 16 24
A1 25 54
A1 52 57
A1 59 62
A1 64 84
A1 74 84
A2 15 20
A2 19 50
OUTPUT
A1 0 14
A1 16 57
A1 59 62
A1 64 84
A2 15 50

Answer 2

You need to have some variables describing currently-accumulated contiguous region. For each line of input, flush the previously-accumulated region if the new input is a new column1 label, or is same label but non-contiguous, or is end-of-file. If it's same label and contiguous yo update the min and max values.
This assumes that columns 1 and 2 are sorted.
The rest is left as an exercise for the reader.