feed_searcher 0.0.5 → 0.0.6

data/README.md

@@ -13,8 +13,12 @@ $ gem install feed_searcher
 ## Usage
 ```ruby
 require "feed_searcher"
-FeedSearcher.search("https://github.com/r7kamura/feed_searcher")
-#=> ["https://github.com/r7kamura/feed_searcher/commits/master.atom"]
+
+FeedSearcher.search("https://github.com/fastladder/feed_searcher")
+#=> ["https://github.com/fastladder/feed_searcher/commits/master.atom"]
+
+FeedSearcher.search("https://github.com/fastladder/feed_searcher/commits/master.atom")
+#=> ["https://github.com/fastladder/feed_searcher/commits/master.atom"]
 ```
 
 
@@ -25,7 +29,7 @@ Let me explain how FeedSearcher works along its execution sequence.
 2. Finds link elements (represented as XPath format)
 3. Extracts URLs from the elements via its `href` attribute
 4. Includes the given URL if its resource itself is a feed
-5. Converts to relative path to absolute path
+5. Converts from relative path to absolute path
 
 FeedSearcher finds link elements matcing following XPath patterns.
 

data/lib/feed_searcher/page.rb

@@ -20,14 +20,14 @@ class FeedSearcher
 
     def feed_urls
       urls = []
-      urls << url if (has_feed_mime_type? || has_feed_extension?) && xml?
+      urls << url if like_xml? && parsable_as_xml? && has_feed_element?
       urls += links.map {|link| link["href"] }
     end
 
     private
 
     def has_xml_declaration?
-      !!body.index("<?xml")
+      !!body.start_with?("<?xml")
     end
 
     def has_feed_mime_type?
@@ -38,12 +38,16 @@ class FeedSearcher
       EXTENSIONS.include?(extension)
     end
 
+    def has_feed_element?
+      root.xpath("contains(' feed RDF rss ', concat(' ', local-name(/*), ' '))")
+    end
+
     def parsable_as_xml?
       !!xml
     end
 
-    def xml?
-      has_xml_declaration? && parsable_as_xml?
+    def like_xml?
+      has_xml_declaration? || has_feed_mime_type? || has_feed_extension?
     end
 
     def url

data/lib/feed_searcher/version.rb

@@ -1,3 +1,3 @@
 class FeedSearcher
-  VERSION = "0.0.5"
+  VERSION = "0.0.6"
 end

data/spec/feed_searcher_spec.rb

@@ -3,7 +3,7 @@ require "active_support/core_ext/string/strip"
 
 describe FeedSearcher do
   describe ".search" do
-    context "when the specified resource is HTML" do
+    context "when there are link elements of feeds in the resource" do
       before do
         stub_request(:get, "http://example.com/").to_return(
           :body => <<-EOS.strip_heredoc
@@ -38,7 +38,7 @@ describe FeedSearcher do
       #   * it keeps other domain
       #   * it converts relative path to absolute url
       #
-      it "returns feed URLs from link elements in the specified resource" do
+      it "includes hrefs of them as feed URLs" do
         FeedSearcher.search("http://example.com/").should == %w[
           http://example.com/1
           http://example.com/2
@@ -50,10 +50,36 @@ describe FeedSearcher do
       end
     end
 
-    context "when the specified resource has feed MIME type and be parsable as XML" do
+    context "when the resource has feed MIME type and parsable XML and rss element" do
       before do
         stub_request(:get, "http://example.com/").to_return(
           :headers => { "Content-Type" => "application/rss+xml; charset=UTF-8" },
+          :body    => <<-EOS.strip_heredoc
+            <rss>
+              <channel>
+                <title>title</title>
+                <link>http://exmple.com/</link>
+                <item>
+                  <title>item title</title>
+                  <link>http://example.com/item</link>
+                  <description>item description</description>
+                </item>
+              </channel>
+            </rss>
+          EOS
+        )
+      end
+
+      it "includes the given URL as a feed URL" do
+        FeedSearcher.search("http://example.com/").should == %w[
+          http://example.com/
+        ]
+      end
+    end
+
+    context "when the resource has XML declaration and parsable XML and rss element" do
+      before do
+        stub_request(:get, "http://example.com/").to_return(
           :body    => <<-EOS.strip_heredoc
             <?xml version="1.0" encoding="UTF-8"?>
             <rss>
@@ -71,18 +97,17 @@ describe FeedSearcher do
         )
       end
 
-      it "returns itself as a feed url" do
+      it "includes the given URL as a feed URL" do
         FeedSearcher.search("http://example.com/").should == %w[
           http://example.com/
         ]
       end
     end
 
-    context "when the specified resource has feed extension and be parsable as XML" do
+    context "when the resource has feed extension and parsable XML and feed element" do
       before do
         stub_request(:get, "http://example.com/feed.atom").to_return(
           :body => <<-EOS.strip_heredoc
-            <?xml version="1.0" encoding="UTF-8"?>
             <feed xmlns="http://www.w3.org/2005/Atom">
               <title>title</title>
               <link rel="self" href="http://example.com/1"/>
@@ -101,11 +126,31 @@ describe FeedSearcher do
         )
       end
 
-      it "returns itself as a feed url" do
+      it "includes the given URL as a feed URL" do
         FeedSearcher.search("http://example.com/feed.atom").should == %w[
           http://example.com/feed.atom
         ]
       end
     end
+
+    context "when the resource has XML declaration and parsable XML and no feed element" do
+      before do
+        stub_request(:get, "http://example.com/p3p.xml").to_return(
+          :headers => { "Content-Type" => "application/xhtml+xml" },
+          :body => <<-EOS.strip_heredoc
+            <?xml version="1.0" encoding="UTF-8"?>
+            <META xmlns="http://www.w3.org/2002/01/P3Pv1">
+              <POLICY-REFERENCES>
+              </POLICY-REFERENCES>
+            </META>
+          EOS
+        )
+      end
+
+      it "does not includes the given URL as a feed URL" do
+        FeedSearcher.search("http://example.com/p3p.xml").should == %w[
+        ]
+      end
+    end
   end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: feed_searcher
 version: !ruby/object:Gem::Version
-  version: 0.0.5
+  version: 0.0.6
   prerelease: 
 platform: ruby
 authors:
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-03-19 00:00:00.000000000 Z
+date: 2013-03-20 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: mechanize
@@ -174,9 +174,6 @@ files:
 - lib/feed_searcher/page.rb
 - lib/feed_searcher/version.rb
 - spec/feed_searcher_spec.rb
-- spec/fixtures/example.atom
-- spec/fixtures/example.html
-- spec/fixtures/example.rss
 - spec/spec_helper.rb
 homepage: https://github.com/r7kamura/feed_searcher
 licenses:
@@ -205,8 +202,5 @@ specification_version: 3
 summary: Search RSS feed URLs from the given URL
 test_files:
 - spec/feed_searcher_spec.rb
-- spec/fixtures/example.atom
-- spec/fixtures/example.html
-- spec/fixtures/example.rss
 - spec/spec_helper.rb
 has_rdoc: 

data/spec/fixtures/example.atom

@@ -1,15 +0,0 @@
-<?xml version="1.0" encoding="UTF-8"?>
-<feed xmlns="http://www.w3.org/2005/Atom">
-  <title>title</title>
-  <link rel="self" href="http://example.com/1"/>
-  <link rel="alternate" href="http://example.com/"/>
-  <entry>
-    <title>item title</title>
-    <link rel="alternate" href="http://example.com/"/>
-    <content type="html">
-      <div xmlns="http://www.w3.org/1999/xhtml">
-        <p>item content</p>
-      </div>
-    </content>
-  </entry>
-</feed>

data/spec/fixtures/example.html

@@ -1,18 +0,0 @@
-<!DOCTYPE HTML>
-<html>
-  <head>
-    <meta charset="UTF-8">
-    <link href="http://example.com/1" rel="alternate" type="application/atom+xml" />
-    <link href="http://example.com/2" rel="alternate" type="application/rdf+xml" />
-    <link href="http://example.com/3" rel="alternate" type="application/rss+xml" />
-    <link href="http://example.com/4" rel="alternate" type="application/xml" />
-    <link href="http://example.com/5" rel="resource"  type="application/rss+xml" />
-    <link href="http://www.example.com/6" rel="alternate"  type="application/rss+xml" />
-    <link href="http://other-example.com/7" rel="alternate"  type="application/rss+xml" />
-    <link href="/8" rel="alternate" type="application/rss+xml" />
-  </head>
-  <body>
-    body
-  </body>
- </html>
- 

data/spec/fixtures/example.rss

@@ -1,12 +0,0 @@
-<?xml version="1.0" encoding="UTF-8"?>
-<rss>
-  <channel>
-    <title>title</title>
-    <link>http://exmple.com/</link>
-    <item>
-      <title>item title</title>
-      <link>http://example.com/item</link>
-      <description>item description</description>
-    </item>
-  </channel>
-</rss>