jekyll-algolia 1.1.2 → 1.1.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 51f8221d0fd59da5c41a3a9dd8496dffd6562989
-  data.tar.gz: 56523270302c22f93d2537cfb8a614bf7c263cca
+  metadata.gz: a394d33bddf856fafbb84d345f2a87987583757b
+  data.tar.gz: eaeaa8ff67fdfda25d94e312cdcc18d21510c9a5
 SHA512:
-  metadata.gz: 731e81066460aed7b1459a821f05ef7c4124a2b5cf0ae831bc3bc7cde7c29f9ede433e619a67e8dda111d7008b1bf0268af087cf57476ee2ab23d0e5c8c4c186
-  data.tar.gz: 15d46a745850b81fa7340fd319bd5eb31530457d890a616f8090da9c3fa0e60fccaa913246c6331ecbf00cdcf8c0482f4511cba425d05fbbc3282eba07e9effc
+  metadata.gz: db3edbb0a108e58ceddc7c9c1b6aaafe9747d968b007e4b9a25c58ee7a878035282e8452ee826b57d26a503c247667fccdae991222ffc7d4aa28aa055876bb37
+  data.tar.gz: 24f04bf128f70fe949f8bc27f59e3ac39cb172771dab6365205c0716cc92434e5e4d6d62f46b0c25368031db28d11944f9c065377ef1e92c6af11cf90e84e9c2

data/lib/errors/no_records_found.txt

@@ -10,7 +10,7 @@ I: `files_to_exclude` option. You are currently excluding the following files:
 I:     {files_to_exclude}
 I:
 I: Also double check that your current value for `nodes_to_index` can actually
-I: match something in the page. You are current indexing the following nodes:
+I: match something in the pages. You are current indexing the following nodes:
 I:     {nodes_to_index}
 I:
 I: Note that all the markup that is defined in the layouts won't be available 

data/lib/jekyll-algolia.rb

@@ -17,8 +17,6 @@ module Jekyll
     require 'jekyll/algolia/extractor'
     require 'jekyll/algolia/indexer'
 
-    @config = {}
-
     # Public: Init the Algolia module
     #
     # config - A hash of Jekyll config option (merge of _config.yml options and
@@ -27,9 +25,9 @@ module Jekyll
     # The gist of the plugin works by instanciating a Jekyll site,
     # monkey-patching its `write` method and building it.
     def self.init(config = {})
-      @config = config
       @start_time = Time.now
-      @site = Jekyll::Algolia::Site.new(@config)
+      config = Configurator.init(config).config
+      @site = Jekyll::Algolia::Site.new(config)
 
       exit 1 unless Configurator.assert_valid_credentials
 
@@ -57,14 +55,6 @@ module Jekyll
       @site.process
     end
 
-    # Public: Get access to the Jekyll config
-    #
-    # All other classes will need access to this config, so we make it publicly
-    # accessible
-    def self.config
-      @config
-    end
-
     # Public: Get access to the Jekyll site
     #
     # Tests will need access to the inner Jekyll website so we expose it here

data/lib/jekyll/algolia/configurator.rb

@@ -6,6 +6,8 @@ module Jekyll
     module Configurator
       include Jekyll::Algolia
 
+      @config = {}
+
       # Algolia default values
       ALGOLIA_DEFAULTS = {
         'extensions_to_index' => nil,
@@ -56,13 +58,37 @@ module Jekyll
         }
       }.freeze
 
+      # Public: Init the configurator with the Jekyll config
+      #
+      # config - The config passed by the `jekyll algolia` command. Default to
+      # the default Jekyll config
+      def self.init(config = nil)
+        # Use the default Jekyll configuration if none specified. Silence the
+        # warning about no config set
+        Logger.silent { config = Jekyll.configuration } if config.nil?
+
+        @config = config
+        @config['exclude'] = files_excluded_from_render
+
+        @config = disable_other_plugins(@config)
+
+        self
+      end
+
+      # Public: Access to the global configuration object
+      #
+      # This is a method around @config so we can mock it in the tests
+      def self.config
+        @config
+      end
+
       # Public: Get the value of a specific Jekyll configuration option
       #
       # key - Key to read
       #
       # Returns the value of this configuration option, nil otherwise
       def self.get(key)
-        Jekyll::Algolia.config[key]
+        config[key]
       end
 
       # Public: Get the value of a specific Algolia configuration option, or
@@ -149,7 +175,8 @@ module Jekyll
       # Markdown files can have many different extensions. We keep the one
       # defined in the Jekyll config
       def self.default_extensions_to_index
-        ['html'] + get('markdown_ext').split(',')
+        markdown_ext = get('markdown_ext') || ''
+        ['html'] + markdown_ext.split(',')
       end
 
       # Public: Setting a default value to ignore index.html/index.md files in
@@ -185,6 +212,37 @@ module Jekyll
         false
       end
 
+      # Public: List of files to exclude from the Jekyll build
+      #
+      # We skip all files usually ignored by Jekyll, plus any file that should
+      # not be indexed.
+      def self.files_excluded_from_render
+        site_exclude = get('exclude') || []
+        algolia_exclude = algolia('files_to_exclude') || []
+
+        excluded_files = site_exclude + algolia_exclude
+
+        # 404 pages are not Jekyll defaults but a convention adopted by GitHub
+        # pages. We don't want to index those.
+        # https://help.github.com/articles/creating-a-custom-404-page-for-your-github-pages-site/
+        excluded_files << '404.html'
+        excluded_files << '404.md'
+
+        excluded_files
+      end
+
+      # Public: Disable features from other Jekyll plugins that might interfere
+      # with the indexing
+      def self.disable_other_plugins(config)
+        # Disable pagination from jekyll-paginate
+        # It creates a lot of /page2/index.html files that are not relevant to
+        # indexing
+        # https://github.com/jekyll/jekyll-paginate/blob/master/lib/jekyll-paginate/pager.rb#L23
+        config['paginate'] = nil
+
+        config
+      end
+
       # Public: Check for any deprecated config option and warn the user
       def self.warn_of_deprecated_options
         # indexing_mode is no longer used

data/lib/jekyll/algolia/file_browser.rb

@@ -34,7 +34,7 @@ module Jekyll
         pathname = Pathname.new(file.path)
         return pathname.cleanpath.to_s if pathname.absolute?
 
-        File.join(Configurator.get('source'), file.path)
+        File.expand_path(File.join(Configurator.get('source'), file.path))
       end
 
       # Public: Return the path of a Jekyll file relative to the Jekyll source
@@ -48,45 +48,12 @@ module Jekyll
         pathname = Pathname.new(file.path)
         return file.path if pathname.relative?
 
-        jekyll_source = Pathname.new(Configurator.get('source'))
+        jekyll_source = Pathname.new(
+          File.expand_path(Configurator.get('source'))
+        )
         pathname.relative_path_from(jekyll_source).cleanpath.to_s
       end
 
-      # Public: Check if the file is a 404 error page
-      #
-      # file - The Jekyll file
-      #
-      # 404 pages are not Jekyll defaults but a convention adopted by GitHub
-      # pages. We don't want to index those.
-      # Source: https://help.github.com/articles/creating-a-custom-404-page-for-your-github-pages-site/
-      #
-      # rubocop:disable Naming/PredicateName
-      def self.is_404?(file)
-        File.basename(file.path, File.extname(file.path)) == '404'
-      end
-      # rubocop:enable Naming/PredicateName
-
-      # Public: Check if the page is a pagination page
-      #
-      # file - The Jekyll file
-      #
-      # `jekyll-paginate` automatically creates pages to paginate through posts.
-      # We don't want to index those
-      def self.pagination_page?(file)
-        # paginate_path contains a special `:num` part that is the page number
-        # We convert that to a regexp
-        paginate_path = Configurator.get('paginate_path')
-        paginate_path_as_regexp = paginate_path.gsub(':num', '([0-9]*)')
-
-        regexp = %r{#{paginate_path_as_regexp}/index\.html$}
-
-        # Make sure all file paths start with a / for comparison
-        filepath = file.path
-        filepath = "/#{filepath}" unless filepath[0] == '/'
-
-        Utils.match?(filepath, regexp)
-      end
-
       # Public: Check if the file has one of the allowed extensions
       #
       # file - The Jekyll file
@@ -101,16 +68,6 @@ module Jekyll
         extensions.include?(extname)
       end
 
-      # Public: Check if the file has been excluded by the user
-      #
-      # file - The Jekyll file
-      #
-      # Files can be excluded either by setting the `files_to_exclude` option,
-      # or by defining a custom hook
-      def self.excluded_by_user?(file)
-        excluded_from_config?(file) || excluded_from_hook?(file)
-      end
-
       # Public: Check if the file has been excluded by `files_to_exclude`
       #
       # file - The Jekyll file
@@ -147,10 +104,8 @@ module Jekyll
       # all the static assets, only keep the actual content.
       def self.indexable?(file)
         return false if static_file?(file)
-        return false if is_404?(file)
-        return false if pagination_page?(file)
         return false unless allowed_extension?(file)
-        return false if excluded_by_user?(file)
+        return false if excluded_from_hook?(file)
 
         true
       end
@@ -196,17 +151,10 @@ module Jekyll
         data.each_key do |key|
           data.delete(key) if respond_to?(key)
         end
-
-        # Excerpt needs a special handling
         data.delete('excerpt')
 
-        # jekyll-asciidoc adds a document key to each page. Once converted to
-        # a string representation, it contains a random number, causing the
-        # object hash to change on each run. This causes the diff method to
-        # fail, so we discard this key.
-        if Utils.instance_of?(data['document'], 'Asciidoctor::Document')
-          data.delete('document')
-        end
+        # Convert all values to a version that can be serialized to JSON
+        data = Utils.jsonify(data)
 
         # Convert all keys to symbols
         data = Utils.keys_to_symbols(data)

data/lib/jekyll/algolia/utils.rb

@@ -88,6 +88,39 @@ module Jekyll
           item[key] == value
         end
       end
+
+      # Public: Convert an object into an object that can easily be converted to
+      # JSON, to be stored as a record
+      #
+      # item - The object to convert
+      #
+      # It will keep any string, number, boolean,boolean,array or nested object,
+      # but will try to stringify other objects, excluding the one that contain
+      # a unique identifier once serialized.
+      def self.jsonify(item)
+        simple_types = [
+          NilClass,
+          TrueClass, FalseClass,
+          Integer, Float,
+          String
+        ]
+        return item if simple_types.member?(item.class)
+
+        # Recursive types
+        return item.map { |value| jsonify(value) } if item.is_a?(Array)
+        if item.is_a?(Hash)
+          return item.map { |key, value| [key, jsonify(value)] }.to_h
+        end
+
+        # Can't be stringified, discard it
+        return nil unless item.respond_to?(:to_s)
+
+        # Discard also if is a serialized version with unique identifier
+        stringified = item.to_s
+        return nil if match?(stringified, /#<[^ ].*@[0-9]* .*>/)
+
+        stringified
+      end
     end
   end
 end

data/lib/jekyll/algolia/version.rb

@@ -2,6 +2,6 @@
 
 module Jekyll
   module Algolia
-    VERSION = '1.1.2'
+    VERSION = '1.1.3'
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: jekyll-algolia
 version: !ruby/object:Gem::Version
-  version: 1.1.2
+  version: 1.1.3
 platform: ruby
 authors:
 - Tim Carry
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2018-02-15 00:00:00.000000000 Z
+date: 2018-02-28 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: algolia_html_extractor
@@ -150,6 +150,20 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '4.3'
+- !ruby/object:Gem::Dependency
+  name: guard
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.14'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.14'
 - !ruby/object:Gem::Dependency
   name: guard-rspec
   requirement: !ruby/object:Gem::Requirement