xmltv-util 0.5.67-0.1 / usr / bin / tv_grab_il

#!/usr/bin/perl -w

eval 'exec /usr/bin/perl -w -S $0 ${1+"$@"}'
    if 0; # not running under some shell

=pod

=head1 NAME

tv_grab_il - Grab TV listings for Israel.

=head1 SYNOPSIS

tv_grab_il --help
  
tv_grab_il --version

tv_grab_il --capabilities

tv_grab_il --description


tv_grab_il [--config-file FILE]
           [--days N] [--offset N] [--slow]
           [--output FILE] [--quiet] [--debug]

tv_grab_il --configure [--config-file FILE]

tv_grab_il --configure-api [--stage NAME]
           [--config-file FILE]
           [--output FILE]

tv_grab_il --list-channels [--config-file FILE]
           [--output FILE] [--quiet] [--debug]

=head1 DESCRIPTION

Output TV listings in XMLTV format for many channels available in Israel.
The data comes from tv-guide.walla.co.il.

First you must run B<tv_grab_il --configure> to choose which channels
you want to receive.

Then running B<tv_grab_il> with no arguments will get a listings in XML
format for the channels you chose for available days including today.

=head1 OPTIONS

B<--configure> Prompt for which channels to download and write the
configuration file.

B<--config-file FILE> Set the name of the configuration file, the
default is B<~/.xmltv/tv_grab_il.conf>.  This is the file written by
B<--configure> and read when grabbing.

B<--output FILE> When grabbing, write output to FILE rather than
standard output.

B<--days N> When grabbing, grab N days rather than all available days.

B<--offset N> Start grabbing at today + N days.  N may be negative.

B<--slow> Get programme descriptions as well as title.  (Will take a
long time with lots of channels selected.)

B<--quiet> Suppress the progress-bar normally shown on standard error.

B<--debug> Provide more information on progress to stderr to help in
debugging.

B<--list-channels> Write output giving <channel> elements for every
channel available (ignoring the config file), but no programmes.

B<--capabilities> Show which capabilities the grabber supports. For more
information, see L<http://wiki.xmltv.org/index.php/XmltvCapabilities>

B<--version> Show the version of the grabber.

B<--help> Print a help message and exit.

=head1 ERROR HANDLING

If the grabber fails to download data for some channel on a specific day,
it will print an errormessage to STDERR and then continue with the other
channels and days. The grabber will exit with a status code of 1 to indicate
that the data is incomplete.

=head1 ENVIRONMENT VARIABLES

The environment variable HOME can be set to change where configuration
files are stored. All configuration is stored in $HOME/.xmltv/. On Windows,
it might be necessary to set HOME to a path without spaces in it.

=head1 SUPPORTED CHANNELS

For information on supported channels, see http://tv-guide.walla.co.il/

=head1 AUTHOR

lightpriest. This documentation and parts of the code
based on various other tv_grabbers from the XMLTV-project.

=head1 SEE ALSO

L<xmltv(5)>.

=cut

use strict;
use Encode;
use XMLTV::Options qw/ParseOptions/;
use XMLTV::ProgressBar;
use XMLTV::Configure::Writer;
use XMLTV::Get_nice qw(get_nice_tree);

use POSIX qw(strftime);
use DateTime;


# only used while testing the 'slow' option
##$XMLTV::Get_nice::Delay = 0;
##use HTTP::Cache::Transparent;
##HTTP::Cache::Transparent::init( {
##    BasePath => '/root/.xmltv/cache',
##    NoUpdate => 60*60,			# cache time in seconds
##    MaxAge => 24,				# flush time in hours
##    Verbose => 1,
##} );


my $channel_link_regexp = "\\?w=\\/\\/\\/[0-9]*\\/\\/[A-Za-z]*\\/1";
my $channel_link_id = "\\?w=\\/\\/\\/([0-9]*)\\/\\/[A-Za-z]*\\/1";

# Use XMLTV::Options::ParseOptions to parse the options and take care of the basic capabilities that a tv_grabber should
my ($opt, $conf) = ParseOptions({
  grabber_name => "tv_grab_il",
  version => '$Id: tv_grab_il,v 1.28 2015/06/28 08:40:31 knowledgejunkie Exp $',
  capabilities => [qw/baseline manualconfig apiconfig/],
  stage_sub => \&config_stage,
  listchannels_sub => \&write_channels,
  description => "Israel (tv-guide.walla.co.il)",
  extra_options    => [qw/slow/],      # grab descriptions from sub-page
});

sub config_stage {
  my ($stage, $conf) = @_;

  die "Unknown stage $stage" unless $stage eq "start";

  my $result;
  my $writer = new XMLTV::Configure::Writer(OUTPUT => \$result, encoding => 'utf-8');
  $writer->start({'generator-info-name' => 'tv_grab_il'});
  $writer->end('select-channels');
  return $result;
}

sub fetch_channels {
  my ($opt, $conf) = @_;

  my $channels = {};

  my $bar = new XMLTV::ProgressBar({
    name => "Fetching channels",
    count => 1
  }) unless ($opt->{quiet} || $opt->{debug});

  # Get the page containing the list of channels
  my $tree = XMLTV::Get_nice::get_nice_tree('http://tv-guide.walla.co.il', undef, 'windows-1255');
  my @channels = $tree->look_down("_tag", "a",
    "href", qr/$channel_link_regexp/,
    sub { !$_[0]->look_down('_tag', 'img') }
  );

  $bar->update() && $bar->finish && undef $bar if defined $bar;

  $bar = new XMLTV::ProgressBar({
    name => "Parsing result",
    count => scalar @channels
  }) unless ($opt->{quiet} || $opt->{debug});

  # Browse through the downloaded list of channels and map them to a hash XMLTV::Writer would understand
  foreach my $channel (@channels) {
    if ($channel->as_text()) {
      my ($id) = $channel->attr('href') =~ /$channel_link_id/;

      # Try to fetch the icon
      my $icon = $channel->parent();
      $icon = $icon->right if $icon;
      $icon = $icon->look_down('_tag', 'a', 'href', qr/$channel_link_regexp/) if $icon;
      $icon = $icon->look_down('_tag', 'img') if $icon;
      $icon = $icon->attr('src') if $icon;

      $channels->{"$id.tv-guide.walla.co.il"} = {
        id => "$id.tv-guide.walla.co.il",
        'display-name' => [[ encode( 'utf-8', $channel->as_text()) ]],
        url => [ $channel->attr('href') ]
      };
      $channels->{"$id.tv-guide.walla.co.il"}->{icon} = [ {src => ($icon || '')} ] if ($icon)

    }

    $bar->update() if defined $bar;
  }

  $bar->finish() && undef $bar if defined $bar;

  # Notifying the user :)
  $bar = new XMLTV::ProgressBar({
    name => "Reformatting",
    count => 1
  }) unless ($opt->{quiet} || $opt->{debug});

  $bar->update() && $bar->finish() if defined $bar;

  return $channels;
}

sub write_channels {
  my $channels = fetch_channels($opt, $conf);

  # Let XMLTV::Writer format the results as a valid xmltv file
  my $result;
  my $writer = new XMLTV::Writer(OUTPUT => \$result, encoding => 'utf-8');
  $writer->start({'generator-info-name' => 'tv_grab_il'});
  $writer->write_channels($channels);
  $writer->end();

  return $result;
}

# Fetch the channels again to see what's available
my $channels = fetch_channels($opt, $conf);

# Configure initial elements for XMLTV::Writer
#
# Create a new hash for the channels so that channels without programmes
# won't appear in the final XML
my $encoding   = 'UTF-8';
my $credits    = {'generator-info-name' => 'tv_grab_il'};
my $w_channels = {};
my $programmes = [];

# Progress Bar :)
my $bar = new XMLTV::ProgressBar({
  name => "Fetching channels listings",
  count => (scalar @{$conf->{channel}}) * $opt->{days}
}) unless ($opt->{quiet} || $opt->{debug});

# Fetch listings per channel
foreach my $channel_id (@{$conf->{channel}}) {

  # Check each channel still exists in walla's channels page
  if ($channels->{$channel_id}) {
    my ($walla_id) = ($channel_id =~ /^([0-9]*)\..*$/);

    # Now grab listings for each channel on each day, according to the options in $opt
    for (my $i=$opt->{offset}; $i < ($opt->{offset} + $opt->{days}); $i++) {
      my $theday = DateTime->today()->add (days => $i)->set_time_zone('Asia/Jerusalem');
      my $url = "http://tv-guide.walla.co.il/?w=/4//$walla_id//" . $theday->day_name() . "/1";

      my $tree = XMLTV::Get_nice::get_nice_tree($url, undef, 'windows-1255');

      if ($tree) {
        my @shows = $tree->look_down('_tag', 'table', 'width', '100%', 'dir', 'ltr', 'cellpadding', '2', 'border', '0');
        if (@shows) {
          foreach my $show (@shows) {
            my $title = $show->look_down('_tag', 'a', 'class', 'w3b');
            my $show_hour_element = $show->look_down('_tag', 'span', 'class', 'w3b txt-w');

            my @show_hour = split(/:/, $show_hour_element->as_text());
            my $show_time = $theday->clone();
            if ($show_hour[0] < 6) {
              $show_time->add (days => 1);
            }
            $show_time->set(hour => $show_hour[0], minute => $show_hour[1], second => 0);


            # If user wants descriptions then get them from the sub-page
            #  (can also get stop time while we're there!)
            my ($start, $stop, $desc) = ('', '', ''); my $stop_time;
            if ( defined $opt->{slow} && $opt->{slow} ) {

              if ( my $detailtree = XMLTV::Get_nice::get_nice_tree('http://tv-guide.walla.co.il/'.$title->attr('href'), undef, 'windows-1255') ) {
                if ( my $table = $detailtree->look_down('_tag', 'table', 'class', 'wp-0-b') ) {   # get 1st "wp-0-b" table
                  if ( my $td = $table->look_down('_tag', 'td', 'class', 'w2b', sub{ !$_[0]->look_down('_tag', 'h2') } ) ) {
                    ($start, $stop) = $td->as_text() =~ /(\d\d:\d\d).*(\d\d:\d\d)/;
                  }
                  if ( my $td = $table->look_down('_tag', 'td', 'class', 'w3') ) {
                    $desc = $td->as_text();
                  }
                }
              }
              if ( $stop ne '' ) {
                my @stop_hour = split(/:/, $stop);
                $stop_time = $show_time->clone();
                $stop_time->add (days => 1)  if ( $stop_hour[0] < $show_hour[0] );  # assumes prog not last > 24hours!
                $stop_time->set(hour => $stop_hour[0], minute => $stop_hour[1], second => 0);
              }

            }

            my $prog= {
              start => $show_time->strftime("%Y%m%d%H%M%S %z"),
              title => [[ encode( 'utf-8', $title->as_text()) ]],
              channel => $channel_id
            };
            $prog->{'stop'} = $stop_time->strftime("%Y%m%d%H%M%S %z")  if defined $stop_time;
            $prog->{'desc'} = [[ encode( 'utf-8', $desc) ]]  if $desc ne '';
            push @{$programmes}, $prog;

          }

          # Add this channel to the finalized XML
          $w_channels->{$channel_id} = $channels->{$channel_id} unless $w_channels->{$channel_id};

        } else {
        }
      } else {
      }

      $bar->update if defined $bar;
    }
  }
}

$bar->finish() && undef $bar if defined $bar;

my %w_args;

if (($opt->{offset} != 0) || ($opt->{days} != -999)) {
  $w_args{offset} = $opt->{offset};
  $w_args{days} = ($opt->{days} == -999) ? 100 : $opt->{days};
  $w_args{cutoff} = '060000';
}

my $data = [];
$data->[0] = $encoding;
$data->[1] = $credits;
$data->[2] = $w_channels;
$data->[3] = $programmes;

XMLTV::write_data($data, %w_args);