sitemap_generator 3.2.1 → 3.3

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: ./
   specs:
-    sitemap_generator (3.2.1)
+    sitemap_generator (3.3)
       builder
 
 GEM

data/README.md

@@ -69,6 +69,7 @@ Does your website use SitemapGenerator to generate Sitemaps?  Where would you be
 
 ## Changelog
 
+* v3.3: **Support creating sitemaps with no index file**.  A big thank-you to [Eric Hochberger][ehoch] for generously paying for this feature.
 * v3.2.1: Fix syntax error in SitemapGenerator::S3Adapter
 * v3.2: **Support mobile tags**, **SitemapGenerator::S3Adapter** a simple S3 adapter which uses Fog and doesn't require CarrierWave; Remove Ask from the sitemap ping because the service has been shutdown; [Turn off `include_index`][include_index_change] by default; Fix the news XML namespace;  Only include autoplay attribute if present
 * v3.1.1: Bugfix: Groups inherit current adapter
@@ -261,6 +262,12 @@ task :refresh_sitemaps do
 end
 ```
 
+### Sitemaps with no Index File
+
+Sometimes you may not want the sitemap index file to be automatically created, for example when you have a small site with only one sitemap file.  Or you may only want an index file created if you have more than one sitemap file.  Or you may never want the index file to be created.
+
+To handle these cases, take a look at the `create_index` option in the Sitemap Options section below.
+
 ### Upload Sitemaps to a Remote Host
 
 > SitemapGenerator::S3Adapter is a simple S3 adapter which was added in v3.2 which
@@ -629,6 +636,8 @@ The options passed to `group` only apply to the links and sitemaps generated in
 
 The following options are supported:
 
+* `create_index` - Supported values: `true`, `false`, `:auto`.  Default: `true`. Whether to create a sitemap index file.  If `true` an index file is always created regardless of how many sitemap files are generated.  If `false` an index file is never created.  If `:auto` an index file is created only when you have more than one sitemap file (i.e. you have added more than 50,000 - `SitemapGenerator::MAX_SITEMAP_LINKS` - links).
+
 * `default_host` - String.  Required.  **Host including protocol** to use when building a link to add to your sitemap.  For example `http://example.com`.  Calling `add '/home'` would then generate the URL `http://example.com/home` and add that to the sitemap.  You can pass a `:host` option in your call to `add` to override this value on a per-link basis.  For example calling `add '/home', :host => 'https://example.com'` would generate the URL `https://example.com/home`, for that link only.
 
 * `filename` - Symbol.  The **base name for the files** that will be generated.  The default value is `:sitemap`.  This yields sitemaps with names like `sitemap1.xml.gz`, `sitemap2.xml.gz`, `sitemap3.xml.gz` etc, and a sitemap index named `sitemap_index.xml.gz`.  If we now set the value to `:geo` the sitemaps would be named `geo1.xml.gz`, `geo2.xml.gz`, `geo3.xml.gz` etc, and the sitemap index would be named `geo_index.xml.gz`.
@@ -860,6 +869,7 @@ Tested and working on:
 
 ## Thanks (in no particular order)
 
+* [Eric Hochberger][ehoch]
 * [Rodrigo Flores](https://github.com/rodrigoflores) for News sitemaps
 * [Alex Soto](http://github.com/apsoto) for Video sitemaps
 * [Alexadre Bini](http://github.com/alexandrebini) for Image sitemaps
@@ -890,3 +900,4 @@ Copyright (c) 2009 Karl Varga released under the MIT license
 [news_tags]:http://www.google.com/support/news_pub/bin/answer.py?answer=74288
 [remote_hosts]:https://github.com/kjvarga/sitemap_generator/wiki/Generate-Sitemaps-on-read-only-filesystems-like-Heroku
 [include_index_change]:https://github.com/kjvarga/sitemap_generator/issues/70
+[ehoch]:https://github.com/ehoch

data/VERSION

@@ -1 +1 @@
-3.2.1
+3.3

data/lib/sitemap_generator/builder/sitemap_file.rb

@@ -74,6 +74,15 @@ module SitemapGenerator
       #   sitemap_url - a SitemapUrl instance
       #   sitemap, options - a Sitemap instance and options hash
       #   path, options - a path for the URL and options hash
+      #
+      # KJV: We should be using the host from the Location object if no host is
+      # specified in the call to add().  The issue is noticeable when we add links
+      # to a sitemap direct as in the following example:
+      #   ls = SitemapGenerator::LinkSet.new(:default_host => 'http://abc.com')
+      #   ls.sitemap_index.add('/link')
+      # This raises a RuntimeError: Cannot generate a url without a host
+      # Expected: the link added to the sitemap should use the host from its
+      # location object if no host has been specified.
       def add(link, options={})
         raise SitemapGenerator::SitemapFinalizedError if finalized?
 

data/lib/sitemap_generator/builder/sitemap_index_file.rb

@@ -25,11 +25,11 @@ module SitemapGenerator
         @filesize = bytesize(@xml_wrapper_start) + bytesize(@xml_wrapper_end)
       end
 
-      # Finalize sitemaps as they are added to the index
+      # Finalize sitemaps as they are added to the index.
       def add(link, options={})
         if file = link.is_a?(SitemapFile) && link
           @sitemaps_link_count += file.link_count
-          file.finalize!
+          file.finalize! unless file.finalized?
         end
         super(SitemapGenerator::Builder::SitemapIndexUrl.new(link, options))
       end

data/lib/sitemap_generator/link_set.rb

@@ -8,7 +8,7 @@ module SitemapGenerator
     @@new_location_opts = [:filename, :sitemaps_path, :sitemaps_namer]
 
     attr_reader :default_host, :sitemaps_path, :filename
-    attr_accessor :verbose, :yahoo_app_id, :include_root, :include_index, :sitemaps_host, :adapter, :yield_sitemap
+    attr_accessor :verbose, :yahoo_app_id, :include_root, :include_index, :sitemaps_host, :adapter, :yield_sitemap, :create_index
 
     # Create a new sitemap index and sitemap files.  Pass a block calls to the following
     # methods:
@@ -99,6 +99,15 @@ module SitemapGenerator
     #
     # * <tt>:verbose</tt> - If +true+, output a summary line for each sitemap and sitemap
     #   index that is created.  Default is +false+.
+    #
+    # * <tt>:create_index</tt> - Supported values: `true`, `false`, `:auto`.  Default: `true`.
+    #   Whether to create a sitemap index file.  If `true` an index file is always created,
+    #   regardless of how many links are in your sitemap.  If `false` an index file is never
+    #   created.  If `:auto` an index file is created only if your sitemap has more than
+    #   50,000 (or SitemapGenerator::MAX_SITEMAP_LINKS) links.
+    #
+    # KJV: When adding a new option be sure to include it in `options_for_group()` if
+    # the option should be inherited by groups.
     def initialize(options={})
       options = SitemapGenerator::Utilities.reverse_merge(options,
         :include_root => true,
@@ -108,7 +117,8 @@ module SitemapGenerator
           :google         => "http://www.google.com/webmasters/sitemaps/ping?sitemap=%s",
           :bing           => "http://www.bing.com/webmaster/ping.aspx?siteMap=%s",
           :sitemap_writer => "http://www.sitemapwriter.com/notify.php?crawler=all&url=%s"
-        }
+        },
+        :create_index => true
       )
       options.each_pair { |k, v| instance_variable_set("@#{k}".to_sym, v) }
 
@@ -356,6 +366,9 @@ module SitemapGenerator
       opts.delete(:public_path)
 
       # Reverse merge the current settings
+      # KJV: This hash could be a problem because it needs to be maintained
+      # when new options are added, but can easily be missed.  We really could
+      # do with a separate SitemapOptions class.
       current_settings = [
         :include_root,
         :include_index,
@@ -364,7 +377,8 @@ module SitemapGenerator
         :sitemaps_host,
         :verbose,
         :default_host,
-        :adapter
+        :adapter,
+        :create_index
       ].inject({}) do |hash, key|
         if value = instance_variable_get(:"@#{key}")
           hash[key] = value
@@ -399,8 +413,12 @@ module SitemapGenerator
     # block passed to create() is empty the default links are still included in the
     # sitemap.
     def finalize_sitemap!
-      add_default_links if !@added_default_links && !@created_group
       return if sitemap.finalized? || sitemap.empty? && @created_group
+      add_default_links if !@added_default_links && !@created_group
+      # This will finalize it.  We add to the index even if not creating an index because
+      # the index keeps track of how many links are in our sitemaps and we need this info
+      # for the summary line.  If not for that problem, I would add the sitemap to
+      # the index only if create_index is truthy.
       add_to_index(sitemap)
       output(sitemap.summary)
     end
@@ -408,7 +426,8 @@ module SitemapGenerator
     # Finalize a sitemap index and output a summary line.  Do nothing if it has already
     # been finalized.
     def finalize_sitemap_index!
-      return if @protect_index || sitemap_index.finalized?
+      return if @protect_index || !@create_index || sitemap_index.finalized?
+      return if @create_index == :auto && sitemap_index.link_count <= 1
       sitemap_index.finalize!
       output(sitemap_index.summary)
     end

data/spec/sitemap_generator/link_set_spec.rb

@@ -20,12 +20,13 @@ describe SitemapGenerator::LinkSet do
     let(:ls) { SitemapGenerator::LinkSet.new }
 
     default_options = {
-      :filename => :sitemap,
+      :filename      => :sitemap,
       :sitemaps_path => nil,
-      :public_path => SitemapGenerator.app.root + 'public/',
-      :default_host => nil,
+      :public_path   => SitemapGenerator.app.root + 'public/',
+      :default_host  => nil,
       :include_index => false,
-      :include_root => true
+      :include_root  => true,
+      :create_index  => true
     }
 
     default_options.each do |option, value|
@@ -369,6 +370,19 @@ describe SitemapGenerator::LinkSet do
       end
     end
 
+    describe "create_index" do
+      it "should inherit the value" do
+        ls.group.create_index.should == ls.create_index
+        ls.create_index = :some_value
+        ls.group.create_index.should == :some_value
+      end
+
+      it "should set the value" do
+        group = ls.group(:create_index => :some_value)
+        group.create_index.should == :some_value
+      end
+    end
+
     describe "should share the current sitemap" do
       it "if only default_host is passed" do
         group = ls.group(:default_host => 'http://newhost.com')
@@ -391,11 +405,6 @@ describe SitemapGenerator::LinkSet do
     end
 
     describe "finalizing" do
-      it "should finalize the sitemaps if a block is passed" do
-        @group = ls.group
-        @group.sitemap.finalized?.should be_false
-      end
-
       it "should only finalize the sitemaps if a block is passed" do
         @group = ls.group
         @group.sitemap.finalized?.should be_false
@@ -531,6 +540,11 @@ describe SitemapGenerator::LinkSet do
       ls.create(options)
       options.should == { :filename => 'sitemaptest', :verbose => false }
     end
+
+    it "should set create_index" do
+      ls.create(:create_index => :auto)
+      ls.create_index.should == :auto
+    end
   end
 
   describe "reset!" do
@@ -706,4 +720,63 @@ describe SitemapGenerator::LinkSet do
       end
     end
   end
+
+  describe "create_index" do
+    describe "when false" do
+      let(:ls)  { SitemapGenerator::LinkSet.new(:default_host => default_host, :create_index => false) }
+
+      it "should not finalize the index" do
+        ls.send(:finalize_sitemap_index!)
+        ls.sitemap_index.finalized?.should be_false
+      end
+
+      it "should still add finalized sitemaps to the index (but the index is never finalized)" do
+        ls.expects(:add_to_index).with(ls.sitemap).once
+        ls.send(:finalize_sitemap!)
+      end
+    end
+
+    describe "when true" do
+      let(:ls)  { SitemapGenerator::LinkSet.new(:default_host => default_host, :create_index => true) }
+
+      it "should always finalize the index" do
+        ls.send(:finalize_sitemap_index!)
+        ls.sitemap_index.finalized?.should be_true
+      end
+
+      it "should add finalized sitemaps to the index" do
+        ls.expects(:add_to_index).with(ls.sitemap).once
+        ls.send(:finalize_sitemap!)
+      end
+    end
+
+    describe "when :auto" do
+      let(:ls)  { SitemapGenerator::LinkSet.new(:default_host => default_host, :create_index => :auto) }
+
+      it "should not finalize the index when it is empty" do
+        ls.sitemap_index.empty?.should be_true
+        ls.send(:finalize_sitemap_index!)
+        ls.sitemap_index.finalized?.should be_false
+      end
+
+      it "should add finalized sitemaps to the index" do
+        ls.expects(:add_to_index).with(ls.sitemap).once
+        ls.send(:finalize_sitemap!)
+      end
+
+      it "should not finalize the index when it has only one link" do
+        ls.sitemap_index.add '/test', :host => default_host
+        ls.sitemap_index.empty?.should be_false
+        ls.send(:finalize_sitemap_index!)
+        ls.sitemap_index.finalized?.should be_false
+      end
+
+      it "should finalize the index when it has more than one link" do
+        ls.sitemap_index.add '/test1', :host => default_host
+        ls.sitemap_index.add '/test2', :host => default_host
+        ls.send(:finalize_sitemap_index!)
+        ls.sitemap_index.finalized?.should be_true
+      end
+    end
+  end
 end

data/spec/sitemap_generator/sitemap_generator_spec.rb

@@ -202,6 +202,100 @@ describe "SitemapGenerator" do
     end
   end
 
+  describe "create_index" do
+
+    before :each do
+      clean_sitemap_files_from_rails_app
+    end
+
+    describe "when true" do
+      let(:ls) { SitemapGenerator::LinkSet.new(:include_root => false, :default_host => 'http://example.com', :create_index => true) }
+
+      it "should always create index" do
+        ls.create { }
+        file_should_exist(rails_path('public/sitemap_index.xml.gz'))
+        file_should_exist(rails_path('public/sitemap1.xml.gz'))
+        file_should_not_exist(rails_path('public/sitemap2.xml.gz'))
+      end
+
+      it "should always create index" do
+        with_max_links(1) do
+          ls.create { add('/one') }
+        end
+        file_should_exist(rails_path('public/sitemap_index.xml.gz'))
+        file_should_exist(rails_path('public/sitemap1.xml.gz'))
+        file_should_not_exist(rails_path('public/sitemap2.xml.gz'))
+      end
+
+      it "should always create index" do
+        with_max_links(1) do
+          ls.create { add('/one'); add('/two') }
+        end
+        file_should_exist(rails_path('public/sitemap_index.xml.gz'))
+        file_should_exist(rails_path('public/sitemap1.xml.gz'))
+        file_should_exist(rails_path('public/sitemap2.xml.gz'))
+      end
+    end
+
+    describe "when false" do
+      let(:ls) { SitemapGenerator::LinkSet.new(:include_root => false, :default_host => 'http://example.com', :create_index => false) }
+
+      it "should never create index" do
+        ls.create { }
+        file_should_not_exist(rails_path('public/sitemap_index.xml.gz'))
+        file_should_exist(rails_path('public/sitemap1.xml.gz'))
+        file_should_not_exist(rails_path('public/sitemap2.xml.gz'))
+      end
+
+      it "should never create index" do
+        with_max_links(1) do
+          ls.create { add('/one') }
+        end
+        file_should_not_exist(rails_path('public/sitemap_index.xml.gz'))
+        file_should_exist(rails_path('public/sitemap1.xml.gz'))
+        file_should_not_exist(rails_path('public/sitemap2.xml.gz'))
+      end
+
+      it "should never create index" do
+        with_max_links(1) do
+          ls.create { add('/one'); add('/two') }
+        end
+        file_should_not_exist(rails_path('public/sitemap_index.xml.gz'))
+        file_should_exist(rails_path('public/sitemap1.xml.gz'))
+        file_should_exist(rails_path('public/sitemap2.xml.gz'))
+      end
+    end
+
+    describe "when :auto" do
+      let(:ls) { SitemapGenerator::LinkSet.new(:include_root => false, :default_host => 'http://example.com', :create_index => :auto) }
+
+      it "should not create index if one sitemap file" do
+        ls.create { }
+        file_should_not_exist(rails_path('public/sitemap_index.xml.gz'))
+        file_should_exist(rails_path('public/sitemap1.xml.gz'))
+        file_should_not_exist(rails_path('public/sitemap2.xml.gz'))
+      end
+
+      it "should not create index if one sitemap file" do
+        with_max_links(1) do
+          ls.create { add('/one') }
+        end
+        file_should_not_exist(rails_path('public/sitemap_index.xml.gz'))
+        file_should_exist(rails_path('public/sitemap1.xml.gz'))
+        file_should_not_exist(rails_path('public/sitemap2.xml.gz'))
+      end
+
+      it "should create index if more than one sitemap file" do
+        with_max_links(1) do
+          ls.create { add('/one'); add('/two') }
+        end
+        file_should_exist(rails_path('public/sitemap_index.xml.gz'))
+        file_should_exist(rails_path('public/sitemap1.xml.gz'))
+        file_should_exist(rails_path('public/sitemap2.xml.gz'))
+      end
+    end
+  end
+
   protected
 
   #
@@ -229,7 +323,7 @@ describe "SitemapGenerator" do
 
   # Better would be to just invoke the environment task and use
   # the interpreter.
-  def execute_sitemap_config
-   SitemapGenerator::Interpreter.run
+  def execute_sitemap_config(opts={})
+   SitemapGenerator::Interpreter.run(opts)
   end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: sitemap_generator
 version: !ruby/object:Gem::Version
-  version: 3.2.1
+  version: '3.3'
   prerelease: 
 platform: ruby
 authors:
@@ -10,11 +10,11 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-08-11 00:00:00.000000000Z
+date: 2012-10-03 00:00:00.000000000Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: mocha
-  requirement: &70187786258340 !ruby/object:Gem::Requirement
+  requirement: &70307560374840 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -22,10 +22,10 @@ dependencies:
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *70187786258340
+  version_requirements: *70307560374840
 - !ruby/object:Gem::Dependency
   name: nokogiri
-  requirement: &70187786257900 !ruby/object:Gem::Requirement
+  requirement: &70307560374400 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -33,10 +33,10 @@ dependencies:
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *70187786257900
+  version_requirements: *70307560374400
 - !ruby/object:Gem::Dependency
   name: rspec
-  requirement: &70187786257480 !ruby/object:Gem::Requirement
+  requirement: &70307560373980 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -44,10 +44,10 @@ dependencies:
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *70187786257480
+  version_requirements: *70307560373980
 - !ruby/object:Gem::Dependency
   name: builder
-  requirement: &70187786257060 !ruby/object:Gem::Requirement
+  requirement: &70307560373560 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -55,7 +55,7 @@ dependencies:
         version: '0'
   type: :runtime
   prerelease: false
-  version_requirements: *70187786257060
+  version_requirements: *70307560373560
 description: SitemapGenerator is an XML Sitemap generator written in Ruby with automatic
   Rails integration.  It supports Video, News, Image and Geo sitemaps and includes
   Rake tasks for managing your sitemaps.
@@ -147,7 +147,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: '0'
       segments:
       - 0
-      hash: -4223635251889236113
+      hash: -845779486658727923
 required_rubygems_version: !ruby/object:Gem::Requirement
   none: false
   requirements:
@@ -156,7 +156,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
       segments:
       - 0
-      hash: -4223635251889236113
+      hash: -845779486658727923
 requirements: []
 rubyforge_project: 
 rubygems_version: 1.8.10