kiosks/lib/site.rb

require 'scraperwiki'
require 'petrify'
require 'csv'
require 'json'
require 'rss'


if ENV['KIOSKS_SITEURL']
  SITEURL = ENV['KIOSKS_SITEURL']
else
  puts "KIOSKS_SITEURL environment variable must be set to the base URL of the site without a trailing slash"
  puts "eg https://kiosks.adrianshort.org"
  exit 1
end

class Site
  def self.generate
    # Home page
    summary = ScraperWiki.select("
      authority_name, status, decision, appeal_status, appeal_decision,
        count(*) as applications
      from applications
      group by authority_name, status, decision, appeal_status, appeal_decision
    ")
    
    q = ScraperWiki.select("
      scraped_at
      from applications
      order by scraped_at desc
      limit 1")

    last_updated = DateTime.parse(q[0]['scraped_at'])

    path = '.'
    Petrify.page(path, 'index', \
      { summary: summary, last_updated: last_updated })
    Petrify.csv(path, 'inlink-summary', summary)  

    # Generate a JSON file with all the data
    apps = ScraperWiki.select("* from applications")
    Petrify.file(path, 'data.json', apps.to_json)

    # New applications page
    apps = ScraperWiki.select("* from `applications`
      order by date_received desc limit 60")
    Petrify.page('new-applications', 'new-applications', { apps: apps, title: "New applications" })

    # Latest decisions page
    apps = ScraperWiki.select("* from `applications`
      order by date_decision desc limit 60")
    path = 'decisions'
    Petrify.page(path, 'decisions', { apps: apps, title: "Latest decisions" })
    Petrify.csv(path, 'inlink-decisions', apps)
    
    # Appeals page
    summary = ScraperWiki.select("
      authority_name, status, decision, appeal_status, appeal_decision,
        count(*) as applications
      from applications
      where (appeal_status is not null
      and appeal_status != 'Unknown')
      or status like '%appeal%'
      group by authority_name, appeal_status, appeal_decision
      collate nocase
    ")
    
    apps = ScraperWiki.select("
      * from applications  
      where (appeal_status is not null
      and appeal_status != 'Unknown')
      or status like '%appeal%'
      collate nocase
    ")
    
    path = 'appeals'
    Petrify.page(path, 'appeals', { summary: summary, apps: apps, title: "Appeals" })
    Petrify.csv(path, 'inlink-appeals', apps)  

    # Media page
    stories = CSV.read('media.csv', :headers => true )
    Petrify.page('media', 'media', { stories: stories, title: "Media" })

    feed = RSS::Maker.make("2.0") do |maker|
      maker.channel.title = "InLinkUK kiosks media coverage"
      maker.channel.description = "News and views about Google's UK street kiosk network."
      maker.channel.link = "#{SITEURL}/media/"
      maker.channel.updated = Time.now.to_s

      stories.each do |story|
        maker.items.new_item do |item|
          item.link = story['url']
          item.title = "%s: %s" % [ story['publication'], story['title'] ]
          item.updated = story['publish_date']
          
          if story['authorities']
            links = []
            story['authorities'].split('|').each do |auth|
              auth.strip!
              links << "<a href=%s>%s</a>" % [ SITEURL + authority_url(auth), auth ]
            end
            item.description = links.join(', ')
          end
        end
      end
    end  
    Petrify.file('media', 'index.xml', feed)
      
    # Authority pages
    auths = ScraperWiki.select("distinct(authority_name) as authority_name
      from applications")

    auths.each do |auth|
      summary = ScraperWiki.select("
        status, decision, appeal_status, appeal_decision, count(*) as qty
        from applications
        where authority_name = ?
        group by status, decision, appeal_status, appeal_decision
        ", auth['authority_name'])

      apps = ScraperWiki.select("* from applications where authority_name = ?
        order by date_received desc", auth['authority_name'])

      this_stories = stories.select do |story|
        if story['authorities']
          story['authorities'].match(auth['authority_name'])
        end
      end

      path = ['authorities', slug(auth['authority_name'])]
      Petrify.page(path, 'authority', \
        { apps: apps, auth: auth, summary: summary, stories: this_stories, title: auth['authority_name'] })
      Petrify.csv(path, slug(auth['authority_name']), apps)
      
      # RSS feed for this authority's media stories
      feed = RSS::Maker.make("2.0") do |maker|
        maker.channel.title = "#{auth['authority_name']} InLinkUK kiosks media coverage"
        maker.channel.description = "News and views about Google's UK street kiosk network in #{auth['authority_name']}."
        maker.channel.link = "#{SITEURL}#{authority_url(auth['authority_name'])}"
        maker.channel.updated = Time.now.to_s

        this_stories.each do |story|
          maker.items.new_item do |item|
            item.link = story['url']
            item.title = "%s: %s" % [ story['publication'], story['title'] ]
            item.updated = story['publish_date']
            
            if story['authorities']
              links = []
              story['authorities'].split('|').each do |auth|
                auth.strip!
                links << "<a href=%s>%s</a>" % [ SITEURL + authority_url(auth), auth ]
              end
              item.description = links.join(', ')
            end
          end
        end
      end
      Petrify.file(path, 'media.xml', feed)
    end
  end
end