How do I extract the value from a nested hash with unknown level of nesting? - perl

What is the best way to walk a nested hash structure to get the value when there can be an unknown number of nested levels?
For example the hash could be any of the following or any level of nesting.
my $hash = { 'known' => { 'a' => { 'b' => 'value' } } };
my $hash = { 'known' => { 'a' => { 'b' => { 'c' => 'value' } } } };
my $hash = { 'known' => { 'a' => { 'b' => { 'c' => { 'd' => 'value' } } } } };
The keys a,b,c below could be any value.
I was thinking I could do it with a recursive function that extracts the key and value and the current level, checks to see if the value is a reference to
a hash and if so calls itself, otherwise I have the value?

An easy-to-read approach:
#!/usr/bin/env perl
use strict;
use warnings;
my $hash = { 'known' => { 'a' => { 'b' => 'value' } } };
print get_deep_values($hash);
sub get_deep_values {
my $hash = shift;
if (ref($hash) eq 'HASH') {
get_deep_values( (values %$hash)[0] )
}
else {
return $hash;
}
}
Output:
value

Yes, recursion works,
use v5.16;
use warnings;
my $hash = { 'known' => { 'a' => { 'b' => 'value' } } };
print sub { map { ref() ? __SUB__->(values %$_) : $_ } #_ }->($hash);
output
value

I assume none of the nested hashes have more than a single element?
A simple iteration is all that's necessary
use strict;
use warnings;
use 5.010;
my #hashes = (
{ 'known' => { 'a' => { 'b' => 'value1' } } },
{ 'known' => { 'a' => { 'b' => { 'c' => 'value2' } } } },
{ 'known' => { 'a' => { 'b' => { 'c' => { 'd' => 'value3' } } } } },
);
say get_value($_) for #hashes;
sub get_value {
my ($hash) = #_;
($hash) = values %$hash while ref $hash eq 'HASH';
$hash;
}
output
value1
value2
value3

Related

Perl how to access a value of nested hash without specific some middle keys (wildcard)

I trying to fetch the facebook mobile posts.
$VAR1 = {
"mf_story_key" => "225164133113094",
"page_id" => "102820022014173",
"page_insights" => {
"102820022014173" => {
"actor_id" => "102820022014173",
"page_id" => "102820022014173",
"page_id_type" => "page",
"post_context" => {
"publish_time" => "1641702174",
"story_name" => "EntStatusCreationStory"
},
"psn" => "EntStatusCreationStory",
"role" => 1,
}
},
"story_attachment_style" => "album",
};
$publish_time = $VAR1->{page_insights}{102820022014173}{post_context}{publish_time};
If 102820022014173 is a dynamic value, how do I access the publish_time value without specific it?
You need to get the keys to the page_insights hash and then iterate through them.
use strict;
use warnings;
use 5.010;
my $post = {
"mf_story_key" => "225164133113094",
"page_id" => "102820022014173",
"page_insights" => {
"102820022014173" => {
"actor_id" => "102820022014173",
"page_id" => "102820022014173",
"page_id_type" => "page",
"post_context" => {
"publish_time" => "1641702174",
"story_name" => "EntStatusCreationStory"
},
"psn" => "EntStatusCreationStory",
"role" => 1,
}
},
"story_attachment_style" => "album",
};
my $insights = $post->{page_insights};
my #insight_ids = keys %{$insights};
for my $id ( #insight_ids ) {
say "ID $id was published at ",
$insights->{$id}{post_context}{publish_time};
}
gives
ID 102820022014173 was published at 1641702174
for my $page_insight ( values( %{ $VAR1->{page_insights} } ) ) {
my $publish_time = $page_insight->{post_context}{publish_time};
...
}
If there's always going to exactly one element,
my $page_insight = ( values( %{ $VAR1->{page_insights} } )[0];
my $publish_time = $page_insight->{post_context}{publish_time};
...
(You can combine the two statements if you so desire.)

How to get the field name in Params::Validate in Perl

validate(
#_,
{
foo => {
callbacks => {
'smaller than a breadbox' => sub { shift() < $breadbox },
'green or blue' => sub {
return 1 if $_[0] eq 'green' || $_[0] eq 'blue';
&$failed(‘**fieldname** value is Invalid’);
}
bar => {
callbacks => {
'yellow or red' => sub {
return 1 if $_[0] eq 'yellow ' || $_[0] eq 'red';
&$failed(‘**fieldname** value is Invalid’);
}
}
}
}
);
Params::Validate - In the above code if my validation fails, I am calling the subroutine failed in which I am displaying the error message. I want to get field name from callbacks in params validate for which the validation has failed instead of directly passing the field name in the failed subroutine. Here foo and bar are fieldnames. How can I get the field name?
You could try something like this:
use strict;
use warnings;
use Params::Validate;
my %template =
(
bar => {
'color correct' => sub {
my ($fieldname, $value) = #_[0..1];
return 1 if $value eq 'green' || $value eq 'blue';
failed("'$fieldname': value '$value' is invalid");
}
},
foo => {
'smaller than a breadbox' => sub {
my ($fieldname, $value) = #_[0..1];
return 1 if $value < 20;
failed("'$fieldname' value '$value' is invalid");
}
}
);
func( bar => 'green', foo => 14 );
func( bar => 'yellow', foo => 17 );
sub func {
my %validate = map { $_ => { callbacks => get_callbacks( $_, $template{$_} ) } }
keys %template;
validate( #_, \%validate );
}
sub get_callbacks {
my ( $fname, $callbacks ) = #_;
my %cb;
for my $name (keys %$callbacks ) {
$cb{$name} = sub {
my $fieldname = $fname;
$callbacks->{$name}->( $fieldname, #_ )
};
}
return \%cb;
}
sub failed {
die $_[0];
}
Note: this uses closures to define a persistent fieldname variable. See perlsub for more information.

How to retrieve values from the hash?

How to retrieve value of d in this hash:
'a' => {
'b' => {
'c' => {
'd' => 'e'
}
You've a 3 level deep hash - assuming you've a typo and your hash looks like:
my %hash = (
'a' => {
'b' => {},
'c' => { 'd' => 'e' }
}
);
print $hash{a}{c}{d};
If that bracket is at the end though:
my %hash = ( 'a' => { 'b' => { 'c' => { 'd' => 'e' } } } );
print $hash{a}{b}{c}{d};

Create deep hash mapping in perl

Below is my Code with the Hash
#!/usr/bin/perl
use warnings;
use JSON::PP; # Just 'use JSON;' on most systems
my %name = (
'sl' => {
'fsd' => {
'conf' => {
'ul' => '/sl/fsd/conf/ul',
'si' => '/sl/fsd/conf/si',
'ho1' => '/sl/fsd/conf/ho1'
}
}
},
're' => {
'fsd' => {
'cron' => {
'README' => '/re/fsd/cron/README'
},
'bin' => {
'db' => {
'smart.p_add_tag' => '/re/fsd/bin/db/smart.p_add_tag',
'smart.p_tag_partition' => '/re/fsd/bin/db/smart.p_tag_partition',
'smart.p_add_tag_type' => '/re/fsd/bin/db/smart.p_add_tag_type'
}
},
'doc' => {
'SMART' => '/re/fsd/doc/SMART',
'README' => '/re/fsd/doc/README'
},
'data' => {
'README' => '/re/fsd/data/README'
},
'conf' => {
'al1' => '/re/fsd/conf/al1',
'file' => '/re/fsd/conf/file',
'ho' => '/re/fsd/conf/ho',
'al3' => '/re/fsd/conf/al3',
'hst' => '/re/fsd/conf/hst',
'us' => '/re/fsd/conf/us',
'README' => '/re/fsd/conf/README',
'al2' => '/re/fsd/conf/al2'
}
}
}
);
(my $root) = keys %name;
my %nodes = ();
my %tree = ();
my #queue = ($root);
list_children(\%name, \#queue, \%nodes) while #queue;
my $tree = build_tree($root, \%nodes);
my $json = JSON::PP->new->pretty; # prettify for human consumption
print $json->encode($tree);
sub list_children {
my $adjac = shift;
my $queue = shift;
my $nodes = shift;
my $node = shift #$queue;
my #children = keys %{$adjac->{$node}};
#children = grep { ! exists $nodes->{$_}} #children;
$nodes->{$node} = \#children;
push #$queue, #children;
}
sub build_tree {
my $root = shift;
my $nodes = shift;
my #children;
for my $child (#{$nodes->{$root}}) {
push #children, build_tree($child, $nodes);
}
my %h = ('text' => $root,
'children' => \#children);
return \%h;
}
I'm trying to output JSONified hash, but it is only traversing upto two levels. whereas i need it to traverse all upto the last child node of each parent. Can someone please help to achieve this.
Below is current output
{
"text" : "sl",
"children" : [
{
"text" : "fsd",
"children" : []
}
]
}
Normally, transforming the hash, and then json-ing is not the most efficient idea, because you're going to make one traversal to transform the hash and JSON's going to make one to json-ify it, and JSON is a type of transform of a hash.
However, JSON is usually done with XS, which means that the second traversal is faster, at least. That and JSON behavior is standardized.
use 5.016;
use strict;
use warnings;
use Data::Dumper ();
use JSON;
my $hash
= {
'Foods' => {
'fruits' => {
'orange' => '1',
'apple' => '2',
},
'Vegetables' => {
'tomato' => '3',
'carrot' => '1',
'cabbage' => '2',
}
}
};
sub descend {
my ( $structure, $block ) = #_;
my $res;
while ( my ( $k, $v ) = each %$structure ) {
$block->( $structure, $k, $v );
if ( ref( $v ) eq 'HASH' ) {
$res = descend( $v, $block );
}
}
return $res;
}
my $new = {};
my $curr = $new;
descend( $hash => sub {
my ( $lvl, $k, $v ) = #_;
my $node = { text => $k };
$curr->{children} //= [];
push $curr->{children}, $node;
if ( ref( $v ) eq 'HASH' ) {
$curr = $node;
}
else {
$node->{children} = { text => $v };
}
});
# allow for the root-level special case, and retrieve the first child.
$new = $new->{children}[0];
say Data::Dumper->Dump( [ $new ], [ '$new' ] );
say JSON->new->encode( $new );

How to merge HashRef in Moose attribute writer?

Having a simple example code
use Modern::Perl;
use Data::Dumper;
package My;
use Moose;
use Method::Signatures::Simple;
has 'result' => (
is => 'rw',
isa => 'HashRef',
default => sub{{}},
clearer => 'clear_result'
);
method run {
$self->clear_result; #clearing the result
$self->result( $self->run_part1 );
$self->do_something;
$self->result( $self->run_part3 ); #need merge
}
method do_something {
$self->result( $self->run_part2 ); #need merge
}
method run_part1 { return { aaa => 'aaa' } }
method run_part2 { return { bbb => 'bbb' } }
method run_part3 { return { ccc => 'ccc' } }
package main;
my $p = My->new;
say Dumper $p->run;
the result (ofcourse) is:
$VAR1 = {
'ccc' => 'ccc'
};
I want the result:
$VAR1 = {
'aaa' => 'aaa'
'bbb' => 'bbb'
'ccc' => 'ccc'
};
so, the question is - how to merge the $self->result HashRef when setting it?
Yes, i can add new method add_result like:
method add_result($hr) {
use Hash::Merge::Simple qw(merge);
$self->result( merge $self->result, $hr );
}
and change everywhere in my code the $self->result to $self->add_result but wonder if there is another solution...
has 'result' => (
acccessor => '_result',
isa => 'HashRef',
default => sub{{}},
clearer => 'clear_result'
);
sub result {
my $self = shift;
if (#_) {
my ($hr) = #_;
return $self->_result( ... merged hash ...);
} else {
return $self->_result();
}
}