Re^2: read flat file and build tree-like structure

Close. Keep track of a reference to the reference.

my %hash;

while (<>) {
  chomp;
  my ($key, $val) = split(/=/, $_, 2);
  my @elts = split(/\./, $key);

  my $p = \\%hash;
  $p = \( ${$p}->{$_} ) for @elts;

  $$p = $val;
}

use Data::Dumper;
print Dumper(\%hash);
[download]

Tested.

Comment on Re^2: read flat file and build tree-like structure Download Code

Replies are listed 'Best First'.
Re^3: read flat file and build tree-like structure by manav_gupta (Acolyte) on Nov 19, 2007 at 17:17 UTC
Wow, that's awesome. I'd been wondering for last few hours on the best way of doing it... thank you so much!	[reply]
Re^4: read flat file and build tree-like structure by manav_gupta (Acolyte) on Nov 19, 2007 at 17:22 UTC
Couple of follow-up questions: 1. Could you please explain what the following snippet does? `my $p = \\%hash; $p = \(${$p}->{$_}) for @elts; $$p = $val;` [download] 2. Is there a quick/easy way to change the key of a hash? Apologies if I'm breaking a rule here (double/cross-posting)...	[reply] [d/l]
Re^5: read flat file and build tree-like structure by ikegami (Patriarch) on Nov 19, 2007 at 19:07 UTC
There's no problem with asking follow up questions. 1. Just like `DiveVal(\%hash, map \$_, @elts) = $val;` [download] it does `$hash{ $elts[0] }{ $elts[1] }...{ $elts[n-1] } = $val;` [download] If you have data like `foo.bar=val foo.bar.baz=val` [download] you'd be better off with pc88mxer's updated code. 2. Keys can be added and removed, not changed. However, combining a removal and an addition is easy. `$hash{$new_key} = delete $hash{$old_key};` [download] or `$hash{$new_key} = delete $hash{$old_key} if exists $hash{$old_key};` [download]	[reply] [d/l] [select]