sitemap_generator 2.2.1 → 3.0.0

data/lib/sitemap_generator/interpreter.rb

@@ -21,7 +21,7 @@ module SitemapGenerator
     #
     # All other options are passed to the LinkSet by setting them using accessor methods.
     def initialize(opts={}, &block)
-      opts.reverse_merge!(:link_set => SitemapGenerator::Sitemap)
+      SitemapGenerator::Utilities.reverse_merge!(opts, :link_set => SitemapGenerator::Sitemap)
       @linkset = opts.delete :link_set
       @linkset.send(:set_options, opts)
       eval(&block) if block_given?

data/lib/sitemap_generator/link_set.rb

@@ -32,11 +32,14 @@ module SitemapGenerator
     def create(opts={}, &block)
       reset!
       set_options(opts)
-      start_time = Time.now if @verbose
+      if verbose
+        start_time = Time.now
+        puts "In #{sitemap_index.location.public_path}"
+      end
       interpreter.eval(:yield_sitemap => @yield_sitemap || SitemapGenerator.yield_sitemap?, &block)
       finalize!
-      end_time = Time.now if @verbose
-      puts sitemap_index.stats_summary(:time_taken => end_time - start_time) if @verbose
+      end_time = Time.now if verbose
+      output(sitemap_index.stats_summary(:time_taken => end_time - start_time)) if verbose
       self
     end
 
@@ -96,18 +99,17 @@ module SitemapGenerator
     # * <tt>:verbose</tt> - If +true+, output a summary line for each sitemap and sitemap
     #   index that is created.  Default is +false+.
     def initialize(options={})
-      options.reverse_merge!({
+      SitemapGenerator::Utilities.reverse_merge!(options,
         :include_root => true,
         :include_index => true,
         :filename => :sitemap,
-        :verbose => false,
         :search_engines => {
           :google         => "http://www.google.com/webmasters/sitemaps/ping?sitemap=%s",
           :ask            => "http://submissions.ask.com/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"
         }
-      })
+      )
       options.each_pair { |k, v| instance_variable_set("@#{k}".to_sym, v) }
 
       # If an index is passed in, protect it from modification.
@@ -125,7 +127,7 @@ module SitemapGenerator
     #   host - host for the link, defaults to your <tt>default_host</tt>.
     def add(link, options={})
       add_default_links if !@added_default_links
-      sitemap.add(link, options.reverse_merge!(:host => @default_host))
+      sitemap.add(link, SitemapGenerator::Utilities.reverse_merge!(options, :host => @default_host))
     rescue SitemapGenerator::SitemapFullError
       finalize_sitemap!
       retry
@@ -227,22 +229,24 @@ module SitemapGenerator
     #
     #   SitemapGenerator::Sitemap.ping_search_engines('http://example.com/sitemap_index.xml.gz', :super_engine => 'http://superengine.com/ping?url=%s')
     def ping_search_engines(*args)
-      engines = args.last.is_a?(Hash) ? args.pop : {}
-      index_url = CGI.escape(args.shift || sitemap_index_url)
-
+      require 'cgi/session'
       require 'open-uri'
       require 'timeout'
 
-      puts "\n" if verbose
+      engines = args.last.is_a?(Hash) ? args.pop : {}
+      index_url = CGI.escape(args.shift || sitemap_index_url)
+
+      output("\n")
       search_engines.merge(engines).each do |engine, link|
         link = link % index_url
+        name = Utilities.titleize(engine.to_s)
         begin
           Timeout::timeout(10) {
             open(link)
           }
-          puts "Successful ping of #{engine.to_s.titleize}" if verbose
+          output("Successful ping of #{name}")
         rescue Timeout::Error, StandardError => e
-          puts "Ping failed for #{engine.to_s.titleize}: #{e.inspect} (URL #{link})" if verbose
+          output("Ping failed for #{name}: #{e.inspect} (URL #{link})")
         end
       end
     end
@@ -296,6 +300,16 @@ module SitemapGenerator
       !!@include_root
     end
 
+    # Set verbose on the instance or by setting ENV['VERBOSE'] to true or false.
+    # By default verbose is true.  When running rake tasks, pass the <tt>-s</tt>
+    # option to rake to turn verbose off.
+    def verbose
+      if @verbose.nil?
+        @verbose = SitemapGenerator.verbose.nil? ? true : SitemapGenerator.verbose
+      end
+      @verbose
+    end
+
     protected
 
     # Set each option on this instance using accessor methods.  This will affect
@@ -319,7 +333,7 @@ module SitemapGenerator
     # change the public path.
     def options_for_group(opts)
       opts.delete(:public_path)
-      opts.reverse_merge!(
+      SitemapGenerator::Utilities.reverse_merge!(opts,
         :include_index => false,
         :include_root => false,
         :sitemap_index => sitemap_index
@@ -340,7 +354,7 @@ module SitemapGenerator
         end
         hash
       end
-      opts.reverse_merge!(current_settings)
+      SitemapGenerator::Utilities.reverse_merge!(opts, current_settings)
       opts
     end
 
@@ -371,7 +385,7 @@ module SitemapGenerator
       add_default_links if !@added_default_links && !@created_group
       return if sitemap.finalized? || sitemap.empty? && @created_group
       sitemap_index.add(sitemap)
-      puts sitemap.summary if verbose
+      output(sitemap.summary)
     end
 
     # Finalize a sitemap index and output a summary line.  Do nothing if it has already
@@ -379,7 +393,7 @@ module SitemapGenerator
     def finalize_sitemap_index!
       return if @protect_index || sitemap_index.finalized?
       sitemap_index.finalize!
-      puts sitemap_index.summary if verbose
+      output(sitemap_index.summary)
     end
 
     # Return the interpreter linked to this instance.
@@ -397,6 +411,13 @@ module SitemapGenerator
       @added_default_links = false
     end
 
+    # Write the given string to STDOUT.  Used so that the sitemap config can be
+    # evaluated and some info output to STDOUT in a lazy fasion.
+    def output(string)
+      return unless verbose
+      puts string
+    end
+
     module LocationHelpers
       public
 
@@ -522,7 +543,7 @@ module SitemapGenerator
       # Update the given attribute on the current sitemap index and sitemap file location objects.
       # But don't create the index or sitemap files yet if they are not already created.
       def update_location_info(attribute, value, opts={})
-        opts.reverse_merge!(:include_index => !@protect_index)
+        SitemapGenerator::Utilities.reverse_merge!(opts, :include_index => !@protect_index)
         @sitemap_index.location[attribute] = value if opts[:include_index] && @sitemap_index && !@sitemap_index.finalized?
         @sitemap.location[attribute] = value if @sitemap && !@sitemap.finalized?
       end

data/lib/sitemap_generator/railtie.rb

@@ -1,7 +1,7 @@
 module SitemapGenerator
   class Railtie < Rails::Railtie
     rake_tasks do
-      load File.expand_path('../../../tasks/sitemap_generator_tasks.rake', __FILE__)
+      require File.expand_path('../tasks', __FILE__)
     end
   end
-end
+end

data/lib/sitemap_generator/sitemap_namer.rb

@@ -15,7 +15,7 @@ module SitemapGenerator
     #   :extension - Default: '.xml.gz'. File extension to append.
     #   :start     - Default: 1. Index at which to start counting.
     def initialize(base, options={});
-      @options = options.reverse_merge(
+      @options = SitemapGenerator::Utilities.reverse_merge(options,
         :extension => '.xml.gz',
         :start => 1
       )

data/lib/sitemap_generator/tasks.rb

@@ -1 +1,53 @@
-load File.expand_path(File.join(File.dirname(__FILE__), '../../tasks/sitemap_generator_tasks.rake'))
+# require this file to load the tasks
+require 'rake'
+
+# Require sitemap_generator at runtime.  If we don't do this the ActionView helpers are included
+# before the Rails environment can be loaded by other Rake tasks, which causes problems
+# for those tasks when rendering using ActionView.
+namespace :sitemap do
+  # Require sitemap_generator only.  When installed as a plugin the require will fail, so in
+  # that case, load the environment first.
+  task :require do
+    begin
+      require 'sitemap_generator'
+    rescue LoadError => e
+      if defined?(Rails)
+        Rake::Task['sitemap:require_environment'].invoke
+      else
+        raise e
+      end
+    end
+  end
+
+  # Require sitemap_generator after loading the Rails environment.  We still need the require
+  # in case we are installed as a gem and are setup to not automatically be required.
+  task :require_environment do
+    if defined?(Rails)
+      Rake::Task['environment'].invoke
+    end
+    require 'sitemap_generator'
+  end
+
+  desc "Install a default config/sitemap.rb file"
+  task :install => ['sitemap:require'] do
+    SitemapGenerator::Utilities.install_sitemap_rb(verbose)
+  end
+
+  desc "Delete all Sitemap files in public/ directory"
+  task :clean => ['sitemap:require'] do
+    SitemapGenerator::Utilities.clean_files
+  end
+
+  desc "Generate sitemaps and ping search engines."
+  task :refresh => ['sitemap:create'] do
+    SitemapGenerator::Sitemap.ping_search_engines
+  end
+
+  desc "Generate sitemaps but don't ping search engines."
+  task 'refresh:no_ping' => ['sitemap:create']
+
+  desc "Generate sitemaps but don't ping search engines.  Alias for refresh:no_ping."
+  task :create => ['sitemap:require_environment'] do
+    SitemapGenerator::Interpreter.run(:config_file => ENV["CONFIG_FILE"], :verbose => verbose)
+  end
+end

data/lib/sitemap_generator/utilities.rb

@@ -33,5 +33,111 @@ module SitemapGenerator
       unknown_keys = hash.keys - [valid_keys].flatten
       raise(ArgumentError, "Unknown key(s): #{unknown_keys.join(", ")}") unless unknown_keys.empty?
     end
+
+    # Return a new hash with all keys converted to symbols, as long as
+    # they respond to +to_sym+.
+    def symbolize_keys(hash)
+      symbolize_keys!(hash.dup)
+    end
+
+    # Destructively convert all keys to symbols, as long as they respond
+    # to +to_sym+.
+    def symbolize_keys!(hash)
+      hash.keys.each do |key|
+        hash[(key.to_sym rescue key) || key] = hash.delete(key)
+      end
+      hash
+    end
+
+    # Rounds the float with the specified precision.
+    #
+    #   x = 1.337
+    #   x.round    # => 1
+    #   x.round(1) # => 1.3
+    #   x.round(2) # => 1.34
+    def round(float, precision = nil)
+      if precision
+        magnitude = 10.0 ** precision
+        (float * magnitude).round / magnitude
+      else
+        float.round
+      end
+    end
+
+    # Allows for reverse merging two hashes where the keys in the calling hash take precedence over those
+    # in the <tt>other_hash</tt>. This is particularly useful for initializing an option hash with default values:
+    #
+    #   def setup(options = {})
+    #     options.reverse_merge! :size => 25, :velocity => 10
+    #   end
+    #
+    # Using <tt>merge</tt>, the above example would look as follows:
+    #
+    #   def setup(options = {})
+    #     { :size => 25, :velocity => 10 }.merge(options)
+    #   end
+    #
+    # The default <tt>:size</tt> and <tt>:velocity</tt> are only set if the +options+ hash passed in doesn't already
+    # have the respective key.
+    def reverse_merge(hash, other_hash)
+      other_hash.merge(hash)
+    end
+
+    # Performs the opposite of <tt>merge</tt>, with the keys and values from the first hash taking precedence over the second.
+    # Modifies the receiver in place.
+    def reverse_merge!(hash, other_hash)
+      hash.merge!( other_hash ){|k,o,n| o }
+    end
+
+    # An object is blank if it's false, empty, or a whitespace string.
+    # For example, "", "   ", +nil+, [], and {} are blank.
+    #
+    # This simplifies:
+    #
+    #   if !address.nil? && !address.empty?
+    #
+    # ...to:
+    #
+    #   if !address.blank?
+    def blank?(object)
+      case object
+      when NilClass, FalseClass
+        true
+      when TrueClass, Numeric
+        false
+      when String
+        object !~ /\S/
+      when Hash, Array
+        object.empty?
+      when Object
+        object.respond_to?(:empty?) ? object.empty? : !object
+      end
+    end
+
+    # An object is present if it's not blank.
+    def present?(object)
+      !blank?(object)
+    end
+
+    # Sets $VERBOSE for the duration of the block and back to its original value afterwards.
+    def with_warnings(flag)
+      old_verbose, $VERBOSE = $VERBOSE, flag
+      yield
+    ensure
+      $VERBOSE = old_verbose
+    end
+
+    def titleize(string)
+      string.gsub!(/_/, ' ')
+      string.split(/(\W)/).map(&:capitalize).join
+    end
+
+    def truthy?(value)
+      ['1', 1, 't', 'true', true].include?(value)
+    end
+
+    def falsy?(value)
+      ['0', 0, 'f', 'false', false].include?(value)
+    end
   end
-end
+end

data/lib/tasks/sitemap_generator_tasks.rake

@@ -0,0 +1 @@
+load(File.expand_path(File.join(File.dirname(__FILE__), '../sitemap_generator/tasks.rb')))

data/spec/blueprint.rb

@@ -0,0 +1,15 @@
+require 'machinist/active_record'
+require 'sham'
+
+Sham.title { Time.now.to_i }
+Content.blueprint do
+  title
+end
+
+module Blueprint
+  def self.seed
+    14.times do |i|
+      content = Content.make(:title => "Link #{i}")
+    end
+  end
+end

data/spec/files/sitemap.create.rb

@@ -0,0 +1,12 @@
+SitemapGenerator::Sitemap.default_host = "http://www.example.com"
+
+SitemapGenerator::Sitemap.create do
+  add '/contents', :priority => 0.7, :changefreq => 'daily'
+
+  # add all individual articles
+  (1..10).each do |i|
+    add "/content/#{i}"
+  end
+
+  add "/merchant_path", :host => "https://www.example.com"
+end

data/spec/files/sitemap.deprecated.rb

@@ -0,0 +1,13 @@
+SitemapGenerator::Sitemap.default_host = "http://www.example.com"
+SitemapGenerator::Sitemap.yahoo_app_id = false
+
+SitemapGenerator::Sitemap.add_links do |sitemap|
+  sitemap.add '/contents', :priority => 0.7, :changefreq => 'daily'
+
+  # add all individual articles
+  (1..10).each do |i|
+    sitemap.add "/content/#{i}"
+  end
+
+  sitemap.add "/merchant_path", :host => "https://www.example.com"
+end

data/spec/files/sitemap.groups.rb

@@ -0,0 +1,37 @@
+SitemapGenerator::Sitemap.default_host = "http://www.example.com"
+SitemapGenerator::Sitemap.create(
+    :include_root => true, :include_index => true,
+    :filename => :new_sitemaps, :sitemaps_path => 'fr/') do
+
+  add('/one', :priority => 0.7, :changefreq => 'daily')
+
+  # Test a new location and filename and sitemaps host
+  group(:sitemaps_path => 'en/', :filename => :xxx,
+      :sitemaps_host => "http://newhost.com") do
+
+    add '/two'
+    add '/three'
+  end
+
+  # Test a namer
+  group(:sitemaps_namer => SitemapGenerator::SitemapNamer.new(:abc, :start => 3)) do
+    add '/four'
+    add '/five'
+    add '/six'
+  end
+
+  add '/seven'
+
+  # This should be in a file of its own
+  group(:sitemaps_host => "http://exceptional.com") do
+    add '/eight'
+    add '/nine'
+  end
+
+  add '/ten'
+
+  # This should have no effect.  Already added default links.
+  group(:include_root => true, :include_index => true) {}
+
+  add "/merchant_path", :host => "https://www.merchanthost.com"
+end

data/spec/sitemap_generator/application_spec.rb

@@ -0,0 +1,69 @@
+require 'spec_helper'
+
+describe SitemapGenerator::Application do
+  before :all do
+    SitemapGenerator::Utilities.with_warnings(nil) do
+      Object.const_set(:Rails, Object.new)
+    end
+  end
+
+  after :all do
+    SitemapGenerator::Utilities.with_warnings(nil) do
+      Object.const_set(:Rails, nil)
+    end
+  end
+
+  before :each do
+    @app = SitemapGenerator::Application.new
+  end
+
+  describe "rails3?" do
+    tests = {
+      :nil => false,
+      '2.3.11' => false,
+      '3.0.1' => true,
+      '3.0.11' => true
+    }
+
+    it "should identify the rails version correctly" do
+      tests.each do |version, result|
+        Rails.expects(:version).returns(version)
+        @app.rails3?.should == result
+      end
+    end
+  end
+
+  describe "with Rails" do
+    before :each do
+      @root = '/test'
+      Rails.expects(:root).returns(@root).at_least_once
+    end
+
+    it "should use the Rails.root" do
+      @app.root.should be_a(Pathname)
+      @app.root.to_s.should == @root
+      (@app.root + 'public/').to_s.should == File.join(@root, 'public/')
+    end
+  end
+
+  describe "with no Rails" do
+    before :each do
+      @rails = Rails
+      Object.send(:remove_const, :Rails)
+    end
+
+    after :each do
+      Object::Rails = @rails
+    end
+
+    it "should not be Rails" do
+      @app.rails?.should be_false
+    end
+
+    it "should use the current working directory" do
+      @app.root.should be_a(Pathname)
+      @app.root.to_s.should == Dir.getwd
+      (@app.root + 'public/').to_s.should == File.join(Dir.getwd, 'public/')
+    end
+  end
+end