libbymiller/randomFeedsToHTML.rb

## randomFeedsToHTML.rb
require 'rss'
require 'open-uri'
require 'open_uri_redirections'
require 'cgi'

links = []

# hackparse opml file
file = File.new("subscriptions.opml", "r")
while (line = file.gets)
  if(line.match(/<outline type=\"rss\"/))
    foo = line.gsub(/\s*<outline type=\"rss\" xmlUrl=\"/,"")
    foo = foo.gsub("\"/>","")
    links.push(foo.chomp!)
  end
end
file.close

# get 10 random feeds

things = links.sample(10)

# now get the feeds and get the latest item

html = "<html><head><meta charset=\"utf-8\" /> </head>\n<body>\n"

things.each do |url|
  puts url

  begin
   open(url,{:allow_redirections => :safe}) do |rss|

    begin
      # false here stops it trying to validate: https://www.ruby-forum.com/topic/4412182
      feed = RSS::Parser.parse(rss, false)
      feed_title = nil
      feed_item_title = nil
      feed_item_link = nil
      feed_item_date = nil
      feed_item_description = nil

      begin
        feed_title = feed.channel.title
      rescue=> a
        feed_title = feed.title
        #puts a.backtrace
      end

      # shouldn't need to do this, can't figure out another way, docs are useless
      feed_item = feed_item.to_s.gsub(/<[^>]*>/,"")

      item = feed.items[0]

      begin
        feed_item_link = item.link.href
      rescue => b
        feed_item_link = item.link
        #puts b.backtrace
      end

      begin
        feed_item_title = item.title.to_s
      rescue => c
        feed_item_title = item.title
        #puts c.backtrace
      end

      # remove tags
      feed_item_title = feed_item_title.to_s.gsub(/<[^>]*>/,"")

      begin
        feed_item_date = item.updated
      rescue => d
        #puts d.backtrace
        begin
          feed_item_date = item.pubDate
        rescue => e
          feed_item_date = item.dc_date
          #puts e.backtrace
        end
      end

      begin
        feed_item_description = item.description
      rescue => f
        #puts f.backtrace
        if(item.content && item.content!="")
          feed_item_description = item.content
        else
          feed_item_description = item.summary
        end
      end

      # remove html tags
      feed_item_description = CGI.unescapeHTML(feed_item_description)
      feed_item_description = feed_item_description.gsub(/<[^>]*>/,"")

      # shorten
      feed_item_description = feed_item_description[0,200]+"..."

      html = html + "<h2><a href='#{feed_item_link}'>#{feed_item_title}</a></h2>\n<p>#{feed_item_description}</p>\n<p><i>From <a href='#{url}'>#{feed_title}</a>, #{feed_item_date}</i></p>\n"
    rescue => g
      puts g.backtrace
    end

   end
  rescue => f
     puts f.backtrace # timeout
  end
  sleep(1)
end

html = html + "</body></html>\n"

File.open("#{Dir.pwd}/latest.html", 'w') { |file| file.write(html) }
	require 'rss'
	require 'open-uri'
	require 'open_uri_redirections'
	require 'cgi'

	links = []

	# hackparse opml file
	file = File.new("subscriptions.opml", "r")
	while (line = file.gets)
	if(line.match(/<outline type=\"rss\"/))
	foo = line.gsub(/\s*<outline type=\"rss\" xmlUrl=\"/,"")
	foo = foo.gsub("\"/>","")
	links.push(foo.chomp!)
	end
	end
	file.close

	# get 10 random feeds

	things = links.sample(10)

	# now get the feeds and get the latest item

	html = "<html><head><meta charset=\"utf-8\" /> </head>\n<body>\n"

	things.each do \|url\|
	puts url

	begin
	open(url,{:allow_redirections => :safe}) do \|rss\|

	begin
	# false here stops it trying to validate: https://www.ruby-forum.com/topic/4412182
	feed = RSS::Parser.parse(rss, false)
	feed_title = nil
	feed_item_title = nil
	feed_item_link = nil
	feed_item_date = nil
	feed_item_description = nil

	begin
	feed_title = feed.channel.title
	rescue=> a
	feed_title = feed.title
	#puts a.backtrace
	end

	# shouldn't need to do this, can't figure out another way, docs are useless
	feed_item = feed_item.to_s.gsub(/<[^>]*>/,"")

	item = feed.items[0]

	begin
	feed_item_link = item.link.href
	rescue => b
	feed_item_link = item.link
	#puts b.backtrace
	end

	begin
	feed_item_title = item.title.to_s
	rescue => c
	feed_item_title = item.title
	#puts c.backtrace
	end

	# remove tags
	feed_item_title = feed_item_title.to_s.gsub(/<[^>]*>/,"")

	begin
	feed_item_date = item.updated
	rescue => d
	#puts d.backtrace
	begin
	feed_item_date = item.pubDate
	rescue => e
	feed_item_date = item.dc_date
	#puts e.backtrace
	end
	end

	begin
	feed_item_description = item.description
	rescue => f
	#puts f.backtrace
	if(item.content && item.content!="")
	feed_item_description = item.content
	else
	feed_item_description = item.summary
	end
	end

	# remove html tags
	feed_item_description = CGI.unescapeHTML(feed_item_description)
	feed_item_description = feed_item_description.gsub(/<[^>]*>/,"")

	# shorten
	feed_item_description = feed_item_description[0,200]+"..."

	html = html + "<h2><a href='#{feed_item_link}'>#{feed_item_title}</a></h2>\n<p>#{feed_item_description}</p>\n<p><i>From <a href='#{url}'>#{feed_title}</a>, #{feed_item_date}</i></p>\n"
	rescue => g
	puts g.backtrace
	end

	end
	rescue => f
	puts f.backtrace # timeout
	end
	sleep(1)
	end

	html = html + "</body></html>\n"

	File.open("#{Dir.pwd}/latest.html", 'w') { \|file\| file.write(html) }