Filmnächte-Scraping

Bald sind wieder, ganz hipp, Filmnächte am Elbufer. Leider ist auf http://filmnaechte-am-elbufer.de/ keine sofort verwertbare Information zu finden.

require 'htree'

#
# Saugen
#

#system("wget -O programm.html 'http://filmnaechte-am-elbufer.de/fn.php?idx=20'")


#
# Parsen
#

doc = HTree(File.new('programm.html')).to_rexml
events = []
spans = {}
doc.each_element('/html/body//table[@style=\'width: 488px\']/tr/td//span') { |span|
  text = span.text.to_s
  text.gsub!(/\&nbsp./, ' ')

  spans[span.attributes['class']] = text if text.size > 0

  if span.attributes['class'] == 'progTitle'
    events << spans
    spans = {'progDay'=>spans['progDay'],
             'progTime'=>spans['progTime'],
             'progWeek'=>spans['progWeek']}
  end
}


#
# Ausgabe tabellarisch
#

column_sizes = Hash.new(0)
events.each { |event|
  event.each { |column,cell|
    column_sizes[column] = cell.size if cell.size > column_sizes[column]
  }
}

events.each { |event|
  ptsd = event['progTitleSpecialDay']
  puts %w(progWeek progDay progTime progTitle).collect { |column|
    event[column].ljust(column_sizes[column] + 2)
  }.to_s.strip + (ptsd ? " (#{ptsd.strip})" : "")
}