scrapers/dhmd-veranstaltungen/scrape.rb

#!/usr/bin/env ruby
# coding: utf-8

require 'open-uri'
require 'nokogiri'
require 'erb'

MONTHS = [
  "Jan", "Feb", "Mär", "Apr",
  "Mai", "Jun", "Jul", "Aug",
  "Sep", "Okt", "Nov", "Dez",
]

def fmt_time t
  t.strftime "%Y%m%dT%H%M%S"
end

class Event
  attr_accessor :title, :url
  attr_reader :date, :location, :dtstart, :dtend

  def initialize
    @location = "Hygienemuseum"
  end

  def date=(s)
    if s =~ /(\d{1,2})\. (.+?), (.+?) Uhr/
      year = Time.now.year
      month = MONTHS.index($2) + 1
      year += 1 if month + 1 < Time.now.month
      day = $1.to_i
      case $3
      when /^(\d+)$/
        @dtstart = Time::local year, month, day, $1.to_i, 0, 0
      when /^(\d+):(\d+)$/
        @dtstart = Time::local year, month, day, $1.to_i, $2.to_i, 0
      when /^(\d+)\s*-\s*(\d+)/
        @dtstart = Time::local year, month, day, $1.to_i, 0, 0
        @dtend = Time::local year, month, day, $2.to_i, 0, 0
      when /^(\d+):(\d+)\s*-\s*(\d+):(\d+)/
        @dtstart = Time::local year, month, day, $1.to_i, $2.to_i, 0
        @dtend = Time::local year, month, day, $3.to_i, $4.to_i, 0
      else
        raise "Invalid timestamp: #{$3.inspect}"
      end
      unless @dtend
        @dtend = @dtstart + 3600
      end
    else
      raise "Invalid date: #{s.inspect}"
    end
  end
end

events = []

url = "https://www.dhmd.de/veranstaltungen/kalender/"
doc = Nokogiri::HTML URI.open(url)
doc.css(".event").each do |content|
  date = content.css(".date").text
  title = content.css("h3").text
  href = content.css(".more").attr('href')
  next unless href and date and title

  event_url = URI.join url, href
  next unless event_url.host == "www.dhmd.de"

  begin
    ev = Event::new
    ev.title = title
    ev.url = event_url
    ev.date = date
    events << ev
  rescue
    STDERR.puts "Omitting: #{$!}"
    STDERR.puts $!.backtrace
  end
end

ical = ERB::new <<~EOF
  BEGIN:VCALENDAR
  VERSION:2.0
  METHOD:PUBLISH
  X-WR-TIMEZONE;VALUE=TEXT:Europe/Berlin
  <% events.each  do |ev| %>
  BEGIN:VEVENT
  SUMMARY:<%= ev.title %>
  DTSTART:<%= fmt_time ev.dtstart %>
  DTEND:<%= fmt_time ev.dtend %>
  UID:<%= ev.url %>
  URL:<%= ev.url %>
  LOCATION:<%= ev.location %>
  END:VEVENT
  <% end  %>
  END:VCALENDAR
EOF

puts ical.result