I'm quite new to Perl and I'm trying to build a hash recursively and getting nowhere. I tried searching for tutorials to dynamically build hashes, but all I could find were introductory articles about hashes. I would be grateful if you point me towards the right direction or suggest a nice article/tutorial.
I'm trying to read from a file which has paths in the form of
one/two/three
four
five/six/seven/eight
and I want to build a hash like
VAR = {
one : {
two : {
three : ""
}
}
four : ""
five : {
six : {
seven : {
eight : ""
}
}
}
}
The script I'm using currently is :
my $finalhash = {};
my @input = <>;
sub constructHash {
my ($hashrf, $line) = @_;
@elements = split(/\//, $line);
if(@elements > 1) {
$hashrf->{shift @elements} = constructHash($hashrf->{$elements[0]}, @elements );
} else {
$hashrf->{shift @elements} = "";
}
return $hashrf;
}
foreach $lines (@input) {
$finalhash = constructHash($finalhash, $lines);
}
This is a bit far-fetched, but it works:
It relies on autovivification, which is admittedly a bit advanced for a beginner.
What would probably make any answer to your question a bit twisted is that you ask for empty strings in the leaves, which is of a different "type" than the hashes of the nodes, and requires a different dereferencing operation.
The basics:
Data::Diver
covers this niche so well that people shouldn't reinvent the wheel.I ran your code and found a few problems:
@elements
properly.constructHash()
is a string, but on the recursive call inside, you pass an array of@elements
Try this.
It produces
Remember, Perl hashes aren't ordered.
I've never done something like this, so this approach is likely to be wrong, but well, here's my shot:
EDIT: Fixed!
EDIT2: A (I think) tail-call optimized version, because!