Web::Scraper nested structures & elements only containing spesific data - perl

Question

I have the following code to scrape a form for inputs and get the attributes id and name.
#!/usr/bin/perl
use warnings;
use strict;
use URI;
use Data::Dumper::Simple;
use Web::Scraper;
my $urlToScrape = "http://digitalarkivet.arkivverket.no/finn_kilde";
my $scrap = scraper {
process 'div.listGroup.open > ul.grouped > li.expandable', 'data[]' => scraper {
process 'input', 'id' => '#id', name => '#name';
process 'label', 'label_for' => '#for';
process 'span.listExpander ', 'Text' => 'TEXT';
process 'ul.sublist1', 'sublist[]' => scraper {
process 'input', 'id' => '#id', name => '#name';
process 'label', 'label_for' => '#for';
process 'span', 'label' => 'TEXT';
};
};
};
my $res = $scrap->scrape(URI->new($urlToScrape));
print Dumper($res);
which gives me (shortend $res to fit screen better)
$res = {
'data' => [
{
'label_for' => 'ka0',
'sublist' => [
{
'label' => 'Statlig folketelling',
'label_for' => 'ka0kt0',
'name' => 'kt[]',
'id' => 'ka0kt0'
}
],
'name' => 'ka[]',
'id' => 'ka0>',
'Text' => 'Folketellinger'
},
{
'sublist' => [
{
'label' => 'Manntall',
'name' => 'kt[]',
'label_for' => 'ka1kt0',
'id' => 'ka1kt0'
}
],
'label_for' => 'ka1',
'id' => 'ka1>',
'name' => 'ka[]',
'Text' => 'Manntall'
},
....
{
'label_for' => 'r0',
'sublist' => [
{
'label_for' => 'r0f0',
'id' => 'r0f0',
'name' => 'f[]',
'label' => "01 Østfold"
}
],
'id' => 'r0',
'name' => 'r[]',
'Text' => "Østlandet"
},
{
'Text' => "Sørlandet",
'id' => 'r1',
'sublist' => [
{
'label_for' => 'r1f0',
'name' => 'f[]',
'id' => 'r1f0',
'label' => '09 Aust-Agder'
}
],
'label_for' => 'r1',
'name' => 'r[]'
}
]
};
I' have 2 issues I need to fix. First, I only want to get data for inputs having 'name' = ka[] (at top level).
Second, I only get data for first ul.sublist1 (If you study the page I'm scraping you can see that several "Kildekategori" have subsets of data, which are revealed if expanded/ clicked upon. Putting brackets on Text[] only gets me the sublist textnames, but not their attributes.
I'm thinking I might have to grab data in 2 scrapes instead, since nested values are revealed by id and label_for.

Answer 1

Solved it by scraping three times, foreach "level"
#!/usr/bin/perl
use strict;
use warnings;
use URI;
use Web::Scraper;
use Data::Dumper::Simple;
my %site;
my #res;
my $i;
my $j;
my $label_for;
my #scrape;
$site{'siteID'} = 1;
$site{'url'} = "http://digitalarkivet.arkivverket.no/finn_kilde";
$site{'name'} = "finn_kilde";
open FIL, ">$site{'name'}.csv" or die $!;
my $seperator=";";
$scrape[0] = scraper {
process 'div.listGroup.open > ul.grouped > li.expandable', 'data[]' => scraper {
process 'input',
'id' => '#id',
'value' => '#value',
'type' => '#type',
'name' => '#name';
process 'label', 'label_for' => '#for';
process 'span.listExpander ', 'text' => 'TEXT';
};
};
$scrape[1] = scraper {
process 'ul.sublist1 > li', 'data[]' => scraper {
process 'input',
'id' => '#id',
'value' => '#value',
'type' => '#type',
'name' => '#name';
process 'label', 'label_for' => '#for';
process 'span', 'text' => 'TEXT';
}
};
$scrape[2] = scraper {
process 'ul.sublist2 > li', 'data[]' => scraper {
process 'input',
'id' => '#id',
'value' => '#value',
'type' => '#type',
'name' => '#name';
process 'label', 'label_for' => '#for';
process 'span', 'text' => 'TEXT';
}
};
for $i (0 .. $#scrape){
$res[$i] = $scrape[$i]->scrape(URI->new($site{'url'}));
unless ($i) {
print FIL join($seperator,"label_for","text","name","value","id","type")."\n";
}
for $j (0 .. $#{$res[$i]->{data}}) {
if (defined($res[$i]->{data}[$j]->{label_for})){
$label_for=$res[$i]->{data}[$j]->{label_for};
} else {
$label_for="";
}
if (length($label_for)>0) {
my $name=$res[$i]->{data}[$j]->{name};
my $text=$res[$i]->{data}[$j]->{text};
my $value=$res[$i]->{data}[$j]->{value};
my $id=$res[$i]->{data}[$j]->{id};
my $type=$res[$i]->{data}[$j]->{type};
my #row=($label_for,$text,$name,$value,$id,$type);
print FIL join($seperator,#row);
print FIL "\n";
}
}
sleep(2);
}
close FIL;
print Dumper(\#res);
1;