in reply to Re^2: Merging partially duplicate lines
in thread Merging partially duplicate lines

Here's a database solution showing the flexibility by adding the weighted average

#!perl
use strict;
use DBI;

# create table
my $dbh = create_db('database.sqlite');

# load data
my @files = qw(fileA.txt fileB.txt);  
for my $file (@files){
  load_db($dbh,$file);
}

# report
my $query = 'SELECT A,B,C,D,AVG(E),SUM(F),
             MIN(E),MAX(E),COUNT(*),SUM(E*F)/SUM(F)
             FROM test
             GROUP BY A,B,C,D
             ORDER BY A,B,C,D';
report($dbh,$query);

[download]
sub report {  
  my ($dbh,$sql) = @_;
  my $ar = $dbh->selectall_arrayref($sql);
  print join "\t",qw(A B C D Avg Sum Min Max Count WeightedAvg),"\n";
  for (@$ar){
    print join "\t",@$_,"\n";
  }
}

sub load_db {
  my ($dbh,$filename) = @_;
  open IN,'<',$filename or die "$!";
  
  my $sql = 'INSERT INTO test VALUES (?,?,?,?,?,?)';
  my $sth = $dbh->prepare($sql);
  my $n = 0;
  while (<IN>){
    next unless (/\S/); # skip blank lines
    my @f = split;
    $sth->execute(@f);
    ++$n;
  }
  print "$n records insert from $filename\n";
  close IN;
}

sub create_db {
  my $dbfile = shift;
  unlink($dbfile) if (-e $dbfile);

  my $dbh = DBI->connect(         
    "dbi:SQLite:dbname=$dbfile", 
    "", "", { RaiseError => 1 }       
    ) or die $DBI::errstr;

  my $sql = "CREATE TABLE test (
    A,B INTEGER,C,D,E REAL,F REAL)";

  $dbh->do($sql) or die $DBI::errstr;
  return $dbh;
}

[download]
poj

In Section Seekers of Perl Wisdom