WWW::Mixi::Scraper::Plugin - base class for plugins


WWW-Mixi-Scraper documentation Contained in the WWW-Mixi-Scraper distribution.

Index


Code Index:

NAME

Top

WWW::Mixi::Scraper::Plugin - base class for plugins

SYNOPSIS

Top

    package WWW::Mixi::Scraper::Plugin::<SamplePlugin>

    use strict;
    use warnings;
    use WWW::Mixi::Scraper::Plugin;

    validator {qw( id is_number )};

    sub scrape {
      my ($self, $html) = @_;

      my %scraper;
      $scraper{...} = scraper {
        process '...',
          text => 'TEXT';  # always plain text
        process '...',
          text => 'HTML';  # always HTML
        process '...',
          depends => $self->html_or_text; # HTML or TEXT
        result qw( text depends );
      };

      return $self->post_process($scraper{...}->scrape(\$html));

      return $arrayref;
    }

    1;

DESCRIPTION

Top

This is a base class for WWW::Mixi::Scraper plugins. You don't need to use base qw( WWW::Mixi::Scraper::Plugin ); just use it. This exports Web::Scraper functions and several public and private methods/functions of its own.

METHODS

Top

new

creates an object.

html_or_text

returns the rendering mode of choice ('TEXT' or 'HTML').

parse

gets content from a uri, scrapes, and returns an array (or a hash reference, etc) of data.

build_uri

used internally to build uri from query paramters (and optional hash). If you want to tweak the uri (e.g. to change authority from 'mixi.jp' to 'video.mixi.jp', etc), provide tweak_uri() in your plugin.

get_content

gets content from the uri specified or from an optional html data.

post_process

does some common tasks such as to make sure the result be an array reference, remove some temporary data to extract exact information, objectify link data (and maybe datetime data in the future?).

validator

prepares some simple validator for query parameters, though I'm not sure if this is really useful.

AUTHOR

Top

Kenichi Ishigaki, <ishigaki at cpan.org>

COPYRIGHT AND LICENSE

Top


WWW-Mixi-Scraper documentation Contained in the WWW-Mixi-Scraper distribution.

package WWW::Mixi::Scraper::Plugin;

use strict;
use warnings;
use Web::Scraper;
use String::CamelCase qw( decamelize );
use WWW::Mixi::Scraper::Utils qw( _force_arrayref _uri _datetime );

sub import {
  my $class = shift;
  my $pkg = caller;

  my @subs = qw(
    new parse
    scraper process process_first result
    get_content post_process
    validator build_uri
    html_or_text _extract_name
  );

  no strict   'refs';
  no warnings 'redefine';
  foreach my $sub ( @subs ) {
    *{"$pkg\::$sub"} = *{"$class\::$sub"};
  }
}

sub new {
  my ($class, %options) = @_;

  bless \%options, $class;
}

sub html_or_text { shift->{mode} || 'HTML' }

sub parse {
  my $self = shift;

  my $res = $self->scrape($self->get_content(@_));

  return ( wantarray and ref $res eq 'ARRAY' )
    ? @{ $res || [] }
    : $res;
}

sub get_content {
  my ($self, %options) = @_;

  my $content = delete $options{html};

  unless ( $content ) {
    $content = $self->{mech}->get_content($self->build_uri(%options));
  }
  die "no content" unless $content;

  # XXX: preserve some tags like <br>?
  # $content =~ s/<br(\s[^>]*)?>/\n/g; # at least preserve as a space
  $content =~ s/&nbsp;/ /g;          # as it'd be converted as '?'

  return $content;
}

sub build_uri {
  my ($self, %query) = @_;

  my ($name) = (ref $self) =~ /::(\w+)$/;
  my $path = sprintf '/%s.pl', decamelize($name);
  my $uri = URI->new($path);

  foreach my $key ( keys %query ) {
    if ( $self->_is_valid( $key, $query{$key} ) ) {
      $uri->query_param( $key => $query{$key} );
    }
  }

  $uri = $self->tweak_uri($uri) if $self->can('tweak_uri');

  $self->{uri} = $uri;  # preserve for later use.

  return $uri;
}

sub validator ($) {
  my $hashref = shift;
  my $pkg = caller;

  my %rules;
  foreach my $key ( keys %{ $hashref } ) {
    my $rule = $hashref->{$key};
    if ( $rule eq 'is_number' ) {
      $rules{$key} = sub {
        my $value = shift;
        $value && $value =~ /^\d+$/ ? 1 : 0;
      };
    }
    if ( $rule eq 'is_number_or_all' ) {
      $rules{$key} = sub {
        my $value = shift;
        $value && $value =~ /^(?:\d+|all)$/ ? 1 : 0;
      };
    }
    if ( $rule eq 'is_anything' ) {
      $rules{$key} = sub { 1 };
    }
  }

  no strict   'refs';
  no warnings 'redefine';
  *{"$pkg\::_is_valid"} = sub { return $rules{$_[1]} && $rules{$_[1]}->($_[2]) };
}

sub post_process {
  my ($self, $data, $callback) = @_;

  my $arrayref = _force_arrayref($data);

  foreach my $item ( @{ $arrayref } ) {
    if ( ref $callback eq 'CODE' ) {
      $callback->($item);
    }
    foreach my $key ( keys %{ $item } ) {
      next unless $item->{$key};
      if ( $key =~ /time$/ ) {
        $item->{$key} = _datetime($item->{$key})
      }
      elsif ( $key =~ /(?:link|envelope|image|background|src|icon)$/ ) {
        $item->{$key} = _uri($item->{$key});
      }
      elsif ( $key eq 'images' ) {
        $item->{$key} = _images($item->{$key});
      }
    }
  }

  $arrayref = [ grep { %{ $_ } && !$_->{_delete} } @{ $arrayref } ];

  return $arrayref;
}

sub _images {
  my $item = shift;

  $item = [ $item ] unless ref $item;  # a thumbnail

  my @images;
  foreach my $i ( @{ $item || [] } ) {
    next unless $item;
    push @images, __images($i);
  }
  return \@images;
}

sub __images {
  my $item = shift;
  my ($link, $thumb);
  unless ( ref $item eq 'HASH' ) {
    $link = $thumb = $item;
    $link  =~ s/s\.jpg$/\.jpg/;
    $thumb =~ s/(?:[^s])\.jpg$/s\.jpg/;
  }
  else { 
    $link  = $item->{link} || '';
    $thumb = $item->{thumb_link};

    if ( $link =~ /MM_openBrWindow\(\s*'([^']+)'/ ) { $link = $1; }
  }
  return { link => _uri($link), thumb_link => _uri($thumb) };
}

sub _extract_name {
  my $item = shift;

  return unless defined $item->{string} && defined $item->{subject};

  my $name = substr( delete $item->{string}, length $item->{subject} );
     $name =~ s/^\s*\(//;
     $name =~ s/\)\s*$//;
  $item->{name} = $name;
}

1;

__END__