Fix the situation when a feed is invalid and the custom parser raises…

… an exception, in this case now this error is log and the plugin continues to operate.

Also intoduced some simple test cases for when the rss is perfect, when
is valid but empty and for when is invalid.

Fixes #18

.gitignore

@@ -0,0 +1 @@
+Gemfile.lock

lib/logstash/inputs/rss.rb

@@ -61,46 +61,59 @@ def run(queue)
 
   def handle_response(response, queue)
     body = response.body
-    # @logger.debug("Body", :body => body)
-    # Parse the RSS feed
-    feed = RSS::Parser.parse(body)
-    feed.items.each do |item|
-      # Put each item into an event
-      @logger.debug("Item", :item => item.author)
-      case feed.feed_type
+    begin
+      feed = RSS::Parser.parse(body)
+      feed.items.each do |item|
+        # Put each item into an event
+        @logger.debug("Item", :item => item.author)
+        case feed.feed_type
         when 'rss'
-          @codec.decode(item.description) do |event|
-            event["Feed"] = @url
-            event["published"] = item.pubDate
-            event["title"] = item.title
-            event["link"] = item.link
-            event["author"] = item.author
-            decorate(event)
-            queue << event
-          end
+          handle_rss_response(queue, item)
         when 'atom'
-          if ! item.content.nil?
-            content = item.content.content
-          else
-            content = item.summary.content
-          end
-          @codec.decode(content) do |event|
-            event["Feed"] = @url
-            event["updated"] = item.updated.content
-            event["title"] = item.title.content
-            event["link"] = item.link.href
-            event["author"] = item.author.name.content
-            unless item.published.nil?
-              event["published"] = item.published.content
-            end
-            decorate(event)
-            queue << event
-          end
+          handle_atom_response(queue, item)
+        end
       end
+    rescue RSS::MissingTagError => e
+      @logger.error("Invalid RSS feed", :exception => e)
+    rescue => e
+      @logger.error("Uknown error while parsing the feed", :url => url, :exception => e)
     end
   end
 
   def stop
     Stud.stop!(@run_thread) if @run_thread
   end
+
+  private
+
+  def handle_atom_response(queue, item)
+    if ! item.content.nil?
+      content = item.content.content
+    else
+      content = item.summary.content
+    end
+    @codec.decode(content) do |event|
+      event["Feed"] = @url
+      event["updated"] = item.updated.content
+      event["title"] = item.title.content
+      event["link"] = item.link.href
+      event["author"] = item.author.name.content
+      unless item.published.nil?
+        event["published"] = item.published.content
+      end
+      decorate(event)
+      queue << event
+    end
+  end
+  def handle_rss_response(queue, item)
+    @codec.decode(item.description) do |event|
+      event["Feed"] = @url
+      event["published"] = item.pubDate
+      event["title"] = item.title
+      event["link"] = item.link
+      event["author"] = item.author
+      decorate(event)
+      queue << event
+    end
+  end
 end # class LogStash::Inputs::Exec

spec/fixtures/rss-invalid.xml

@@ -0,0 +1,5 @@
+<?xml version="1.0" encoding="windows-1252"?>
+<rss version="2.0">
+  <channel>
+   </channel>
+</rss>

spec/fixtures/rss-zero-feeditems.xml

@@ -0,0 +1,8 @@
+<?xml version="1.0" encoding="windows-1252"?>
+<rss version="2.0">
+  <channel>
+    <title>Sample Feed - Favorite RSS Related Software &amp; Resources</title>
+    <description>Take a look at some of FeedForAll&apos;s favorite software and resources for learning more about RSS.</description>
+    <link>http://www.feedforall.com</link>
+   </channel>
+</rss>

spec/fixtures/sample-feed.xml

@@ -0,0 +1,52 @@
+<?xml version="1.0" encoding="windows-1252"?>
+<rss version="2.0">
+  <channel>
+    <title>Sample Feed - Favorite RSS Related Software &amp; Resources</title>
+    <description>Take a look at some of FeedForAll&apos;s favorite software and resources for learning more about RSS.</description>
+    <link>http://www.feedforall.com</link>
+    <category domain="www.dmoz.com">Computers/Software/Internet/Site Management/Content Management</category>
+    <copyright>Copyright 2004 NotePage, Inc.</copyright>
+    <docs>http://blogs.law.harvard.edu/tech/rss</docs>
+    <language>en-us</language>
+    <lastBuildDate>Mon, 1 Nov 2004 13:17:17 -0500</lastBuildDate>
+    <managingEditor>marketing@feedforall.com</managingEditor>
+    <pubDate>Tue, 26 Oct 2004 14:06:44 -0500</pubDate>
+    <webMaster>webmaster@feedforall.com</webMaster>
+    <generator>FeedForAll Beta1 (0.0.1.8)</generator>
+    <image>
+      <url>http://www.feedforall.com/feedforall-temp.gif</url>
+      <title>FeedForAll Sample Feed</title>
+      <link>http://www.feedforall.com/industry-solutions.htm</link>
+      <description>FeedForAll Sample Feed</description>
+      <width>144</width>
+      <height>117</height>
+    </image>
+    <item>
+      <title>RSS Resources</title>
+      <description>Be sure to take a look at some of our favorite RSS Resources&lt;br&gt;
+&lt;a href=&quot;http://www.rss-specifications.com&quot;&gt;RSS Specifications&lt;/a&gt;&lt;br&gt;
+&lt;a href=&quot;http://www.blog-connection.com&quot;&gt;Blog Connection&lt;/a&gt;&lt;br&gt;
+&lt;br&gt;</description>
+      <link>http://www.feedforall.com</link>
+      <pubDate>Tue, 26 Oct 2004 14:01:01 -0500</pubDate>
+    </item>
+    <item>
+      <title>Recommended Desktop Feed Reader Software</title>
+      <description>&lt;b&gt;FeedDemon&lt;/b&gt; enables you to quickly read and gather information from hundreds of web sites - without having to visit them. Don&apos;t waste any more time checking your favorite web sites for updates. Instead, use FeedDemon and make them come to you. &lt;br&gt;
+More &lt;a href=&quot;http://store.esellerate.net/a.asp?c=1_SKU5139890208_AFL403073819&quot;&gt;FeedDemon Information&lt;/a&gt;</description>
+      <link>http://www.feedforall.com/feedforall-partners.htm</link>
+      <pubDate>Tue, 26 Oct 2004 14:03:25 -0500</pubDate>
+    </item>
+    <item>
+      <title>Recommended Web Based Feed Reader Software</title>
+      <description>&lt;b&gt;FeedScout&lt;/b&gt; enables you to view RSS/ATOM/RDF feeds from different sites directly in Internet Explorer. You can even set your Home Page to show favorite feeds. Feed Scout is a plug-in for Internet Explorer, so you won&apos;t have to learn anything except for how to press 2 new buttons on Internet Explorer toolbar. &lt;br&gt;
+More &lt;a href=&quot;http://www.bytescout.com/feedscout.html&quot;&gt;Information on FeedScout&lt;/a&gt;&lt;br&gt;
+&lt;br&gt;
+&lt;br&gt;
+&lt;b&gt;SurfPack&lt;/b&gt; can feature search tools, horoscopes, current weather conditions, LiveJournal diaries, humor, web modules and other dynamically updated content. &lt;br&gt;
+More &lt;a href=&quot;http://www.surfpack.com/&quot;&gt;Information on SurfPack&lt;/a&gt;&lt;br&gt;</description>
+      <link>http://www.feedforall.com/feedforall-partners.htm</link>
+      <pubDate>Tue, 26 Oct 2004 14:06:44 -0500</pubDate>
+    </item>
+  </channel>
+</rss>

spec/inputs/rss_spec.rb

@@ -1,6 +1,7 @@
 # encoding: utf-8
 require "logstash/devutils/rspec/spec_helper"
 require "logstash/inputs/rss"
+require 'ostruct'
 
 describe LogStash::Inputs::Rss do
   describe "stopping" do
@@ -11,4 +12,85 @@
     end
     it_behaves_like "an interruptible input plugin"
   end
-end
+
+  describe "fetching input" do
+
+    let(:config) do
+      {
+        "url" => "http://www.example.com/foo.rss",
+        "interval" => 10
+      }
+    end
+
+    let(:sample) do
+      body = File.read(File.join(File.dirname(__FILE__), "..", "fixtures", "sample-feed.xml"))
+      OpenStruct.new(:body => body)
+    end
+
+    before(:each) do
+      allow(Faraday).to receive(:get).with(config["url"]).and_return(sample)
+    end
+
+    let(:data) do
+      plugin = described_class.new(config)
+      plugin_input(plugin) do |queue|
+        sleep 0.1 while queue.empty?
+        events = []
+        queue.size.times { |i| events << queue.pop }
+        events
+      end
+    end
+
+    it "fetchs all items" do
+      expect(data.count).to be > 0
+    end
+
+    context "when the feed is invalid" do
+
+      let(:sample) do
+        body = File.read(File.join(File.dirname(__FILE__), "..", "fixtures", "rss-invalid.xml"))
+        OpenStruct.new(:body => body)
+      end
+
+      let(:plugin) { described_class.new(config) }
+
+      it "fetchs no items and causes no errors" do
+        events = []
+        expect {
+          plugin_input(plugin) do |queue|
+            sleep 1
+            events = []
+            queue.size.times { |i| events << queue.pop }
+            events
+          end
+        }.not_to raise_error
+        expect(events.count).to be == 0
+      end
+    end
+
+
+    context "when the feed is valid, but has zero items" do
+
+      let(:sample) do
+        body = File.read(File.join(File.dirname(__FILE__), "..", "fixtures", "rss-zero-feeditems.xml"))
+        OpenStruct.new(:body => body)
+      end
+
+      let(:plugin) { described_class.new(config) }
+
+      it "fetchs no items and causes no errors" do
+        events = []
+        expect {
+          plugin_input(plugin) do |queue|
+            sleep 1
+            events = []
+            queue.size.times { |i| events << queue.pop }
+            events
+          end
+        }.not_to raise_error
+        expect(events.count).to be == 0
+      end
+    end
+
+  end
+end