A Ruby-script to react to RSS feeds

feedme

#!/usr/bin/env ruby -ws
require 'open-uri'
require 'nokogiri'
require 'yaml'
require 'pathname'
require 'ostruct'
require 'pp'
require 'forwardable'
require 'minitest/unit'


# A FeedMe has an array of Feeds; when it is "call"ed, it "calls" each
# of them in turn. Each Feed is created from a hash of options.
class FeedMe
  def each(&block)
    feeds.each do
      matches.each do
        yield match
      end
    end
  end
  include Enumerable

  def initialize(feeds)
    @feeds = feeds.map { |f| f.symbolize_keys }
  end

  def feeds
    @feeds.map { |f| Feed.new(f) }
  end

  def parser
    @parser ||= parsers[*(super || :rss).to_sym]
  end

  def parse
    instance_eval(&parser)
  end

  def self.parsers(*args)
    @parsers ||= Hash.new { |h, k| h[k] = Module.new }
  end

  class Feed
    extend Forwardable

    def initialize(config)
      @config = config.symbolize_keys
    end

    def xml
      return @xml if instance_variable_defined?(:"@xml")
      @xml = Nokogiri::XML(read) rescue nil
    end

    def method_missing(m, *args, &block)
      return @config[m] if @config.key?(m)
      return xml.send(m, *args, &block) if xml and xml.respond_to?(m)
      super
    end

    def read
      return "" if feed.nil?
      return feed.read if feed.respond_to?(:read)
      open(feed).read rescue feed
    end

    def self.match(s)
      Regexp.new(Time.now.strftime(s))
    end
  end
end

class Hash
  def symbolize_keys
    {}.tap { |r| each { |k, v| r[k.to_sym] = v } }
  end unless instance_methods.include?(:symbolize_keys)

  def stringify_keys
    {}.tap { |r| each { |k, v| r[k.to_s] = v } }
  end unless instance_methods.include?(:stringify_keys)
end

class FeedMeTest < MiniTest::Unit::TestCase
  def setup
  end

  def test_the_matches_method
    m = Time.stub(:now, Time.new(2000)) { FeedMe::Feed.match("={1,2}%Y-*") }
    assert_match m, "=2000"
    refute_match m, "=2001"
    assert_match m, "==2000--"
    refute_match m, "2000"
    assert_match m, "Hello =2000 world"
    refute_match m, "something else"
  end

  def test_creation_of_a_FeedMe
    assert_empty FeedMe.new([]).feeds
  end

  def test_FeedMe_is_enumerable
    fm = FeedMe.new([{x: :a}, {x: :b}, {x: :c}])
    fm.each { |f| assert_kind_of FeedMe::Feed, f }
    assert_kind_of FeedMe::Feed, fm.first
    assert_equal %i[a b c], fm.map(&:x)
  end

  def test_Feed_quacks_like_an_IO
    f = FeedMe::Feed.new(feed: __FILE__)
    assert_equal open(__FILE__).read, f.read
  end

  def test_an_xml_Feed_quacks_like_a_Nokogiri_document
    feed = "<a><b>foo</b></a>"
    f = FeedMe::Feed.new(feed: feed)
    assert_equal 'foo', (f % "b").text
    feed = StringIO.new("<a><b>bar</b></a>")
    f = FeedMe::Feed.new(feed: feed)
    assert_equal 'bar', (f % "b").text
  end

  def test_a_Feed_can_show_its_title
    f = FeedMe::Feed.new(feed: DATA)
  end

  def test_command_line
  end
end

if __FILE__ == $0
  puts "feedme v0.0.1" if defined?($V)
  $verbose = defined?($v)
  require 'minitest/autorun' if defined?($t)
  if defined?($h)
    puts (<<-EOF)
feedme - do stuff with feeds
Usage:
  feedme [config]   run feedme (default config is ~/.feedme/config)
  feedme -v         print the version
  feedme -t         run a test suite
  feedme -h         print this message
    EOF
    exit 0
  end

  ARGV << "~/.feedme/config" if ARGV.empty?
  configs = ARGV.map do |file|
    (YAML.load(File.read(File.expand_path(file))))
  end.flatten
  FeedMe.new(configs)
end

__END__
<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:media="http://search.yahoo.com/mrss/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0">
  <channel> 
    <title>BBC News - Home</title>  
    <link>http://www.bbc.co.uk/news/#sa-ns_mchannel=rss&amp;ns_source=PublicRSS20-sa</link>  
    <description>The latest stories from the Home section of the BBC News web site.</description>  
    <language>en-gb</language>  
    <lastBuildDate>Thu, 21 Nov 2013 18:50:04 GMT</lastBuildDate>  
    <copyright>Copyright: (C) British Broadcasting Corporation, see http://news.bbc.co.uk/2/hi/help/rss/4498287.stm for terms and conditions of reuse.</copyright>  
    <image> 
      <url>http://news.bbcimg.co.uk/nol/shared/img/bbc_news_120x60.gif</url>  
      <title>BBC News - Home</title>  
      <link>http://www.bbc.co.uk/news/#sa-ns_mchannel=rss&amp;ns_source=PublicRSS20-sa</link>  
      <width>120</width>  
      <height>60</height> 
    </image>  
    <ttl>15</ttl>  
    <atom:link href="http://feeds.bbci.co.uk/news/rss.xml" rel="self" type="application/rss+xml"/>  
    <item> 
      <title>Women 'held as slaves for 30 years'</title>  
      <description>Three women are "rescued" from a house in south London as police investigate claims they were held as slaves for 30 years, with one thought to have been in captivity her whole life.</description>  
      <link>http://www.bbc.co.uk/news/uk-england-london-25040741#sa-ns_mchannel=rss&amp;ns_source=PublicRSS20-sa</link>  
      <guid isPermaLink="false">http://www.bbc.co.uk/news/uk-england-london-25040741</guid>  
      <pubDate>Thu, 21 Nov 2013 18:28:14 GMT</pubDate>  
      <media:thumbnail width="66" height="49" url="http://news.bbcimg.co.uk/media/images/71268000/jpg/_71268391_44797439.jpg"/>  
      <media:thumbnail width="144" height="81" url="http://news.bbcimg.co.uk/media/images/71268000/jpg/_71268392_44797439.jpg"/> 
    </item>  
    <item> 
      <title>NI soldiers 'must be prosecuted'</title>  
      <description>Anyone actively involved in shooting unarmed civilians during the Troubles must be prosecuted, Northern Ireland's former deputy first minister says.</description>  
      <link>http://www.bbc.co.uk/news/uk-25037497#sa-ns_mchannel=rss&amp;ns_source=PublicRSS20-sa</link>  
      <guid isPermaLink="false">http://www.bbc.co.uk/news/uk-25037497</guid>  
      <pubDate>Thu, 21 Nov 2013 18:05:41 GMT</pubDate>  
      <media:thumbnail width="66" height="49" url="http://news.bbcimg.co.uk/media/images/71265000/jpg/_71265296_ira624getty.jpg"/>  
      <media:thumbnail width="144" height="81" url="http://news.bbcimg.co.uk/media/images/71265000/jpg/_71265297_ira624getty.jpg"/> 
    </item> 
  </channel>
</rss>