Perl : Need to append two columns if the ID's are repeating - perl

Question

If id gets repeated I am appending app1, app2 and printing it once.
Input:
id|Name|app1|app2
1|abc|234|231|
2|xyz|123|215|
1|abc|265|321|
3|asd|213|235|
Output:
id|Name|app1|app2
1|abc|234,265|231,321|
2|xyz|123|215|
3|asd|213|235|
Output I'm getting:
id|Name|app1|app2
1|abc|234,231|
2|xyz|123,215|
1|abc|265,321|
3|asd|213,235|
My Code:
#! usr/bin/perl
use strict;
use warnings;
my $basedir = 'E:\Perl\Input\\';
my $file ='doctor.txt';
my $counter = 0;
my %RepeatNumber;
my $pos=0;
open(OUTFILE, '>', 'E:\Perl\Output\DoctorOpFile.csv') || die $!;
open(FH, '<', join('', $basedir, $file)) || die $!;
my $line = readline(FH);
unless ($counter) {
chomp $line;
print OUTFILE $line;
print OUTFILE "\n";
}
while ($line = readline(FH)) {
chomp $line;
my #obj = split('\|',$line);
if($RepeatNumber{$obj[0]}++) {
my $str1= join("|",$obj[0]);
my $str2=join(",",$obj[2],$obj[3]);
print OUTFILE join("|",$str1,$str2);
print OUTFILE "\n";
}
}

Answer 1

This should do the trick:
use strict;
use warnings;
my $file_in = "doctor.txt";
open (FF, "<$file_in");
my $temp = <FF>; # remove first line
my %out;
while (<FF>)
{
my ($id, $Name, $app1, $app2) = split /\|/, $_;
$out{$id}[0] = $Name;
push #{$out{$id}[1]}, $app1;
push #{$out{$id}[2]}, $app2;
}
foreach my $key (keys %out)
{
print $key, "|", $out{$key}[0], "|", join (",", #{$out{$key}[1]}), "|", join (",", #{$out{$key}[2]}), "\n";
}
EDIT
To see what the %out contains (in case it's not clear), you can use
use Data::Dumper;
and print it via
print Dumper(%out);

Answer 2

I'd tackle it like this:
#!/usr/bin/env perl
use strict;
use warnings;
use Data::Dumper;
use 5.14.0;
my %stuff;
#extract the header row.
#use the regex to remove the linefeed, because
#we can't chomp it inline like this.
#works since perl 5.14
#otherwise we could just chomp (#header) later.
my ( $id, #header ) = split( /\|/, <DATA> =~ s/\n//r );
while (<DATA>) {
#turn this row into a hash of key-values.
my %row;
( $id, #row{#header} ) = split(/\|/);
#print for diag
print Dumper \%row;
#iterate each key, and insert into $row.
foreach my $key ( keys %row ) {
push( #{ $stuff{$id}{$key} }, $row{$key} );
}
}
#print for diag
print Dumper \%stuff;
print join ("|", "id", #header ),"\n";
#iterate ids in the hash
foreach my $id ( sort keys %stuff ) {
#join this record by '|'.
print join('|',
$id,
#turn inner arrays into comma separated via map.
map {
my %seen;
#use grep to remove dupes - e.g. "abc,abc" -> "abc"
join( ",", grep !$seen{$_}++, #$_ )
} #{ $stuff{$id} }{#header}
),
"\n";
}
__DATA__
id|Name|app1|app2
1|abc|234|231|
2|xyz|123|215|
1|abc|265|321|
3|asd|213|235|
This is perhaps a bit overkill for your application, but it should handle arbitrary column headings and arbitary numbers of duplicates. I'll coalesce them though - so the two abc entries don't end up abc,abc.
Output is:
id|Name|app1|app2
1|abc|234,265|231,321
2|xyz|123|215
3|asd|213|235

Answer 3

Another way of doing it which doesn't use a hash (in case you want to be more memory efficient), my contribution lies under the opens:
#!/usr/bin/perl
use strict;
use warnings;
my $basedir = 'E:\Perl\Input\\';
my $file ='doctor.txt';
open(OUTFILE, '>', 'E:\Perl\Output\DoctorOpFile.csv') || die $!;
select(OUTFILE);
open(FH, '<', join('', $basedir, $file)) || die $!;
print(scalar(<FH>));
my #lastobj = (undef);
foreach my $obj (sort {$a->[0] <=> $b->[0]}
map {chomp;[split('|')]} <FH>) {
if(defined($lastobj[0]) &&
$obj[0] eq $lastobj[0])
{#lastobj = (#obj[0..1],
$lastobj[2].','.$obj[2],
$lastobj[3].','.$obj[3])}
else
{
if($lastobj[0] ne '')
{print(join('|',#lastobj),"|\n")}
#lastobj = #obj[0..3];
}
}
print(join('|',#lastobj),"|\n");
Note that split, without it's third argument ignores empty elements, which is why you have to add the last bar. If you don't do a chomp, you won't need to supply the bar or the trailing hard return, but you would have to record $obj[4].