algoliasearch-jekyll 0.1.2 → 0.1.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 33c6e72889f35db789517b3a9a8b86e97355dac4
-  data.tar.gz: b10d1a46e90b2fe938aebc8f801d3a1b034f029e
+  metadata.gz: e1b192525e42f3f0b61d0587ab11fdef936acc10
+  data.tar.gz: b8557b3ae3e1df1d6e3561c3614a8c4e5e3dc23e
 SHA512:
-  metadata.gz: 1687ec9dcfe975c964ca5d27c3a6f553d746399b1d7db71462bdad3463e40f0cecf89db9624d1eab80b7b07fe9880ecf0ead27aa68db7c74cdf6608ed83607c6
-  data.tar.gz: df50afe1129386a620c1df879b3c90645d49e783e2a6e33fa86c1078ebfb717968f9a1576a99fc59022a612101dd605f1c5cadebc1f5d47640abb45cf7f5d987
+  metadata.gz: b52e88d691848055afbe673401adbef36faf32ad3de1c62985577e0790fd6f97eb4ad8dd0e4f717c6c43b38b5c41a194652e254ed91fdc397931f413fcab3670
+  data.tar.gz: 356ea6edf1d3fdc9ff1b1792294afaa966661ab3c50e3a6eee0d7567e381f742df423b3f6349eb7761393385635c9599144e1e9a0ba2c4c219580790e2eea7f5

data/lib/push.rb

@@ -42,6 +42,10 @@ class AlgoliaSearchJekyllPush < Jekyll::Command
       true
     end
 
+    def excluded_file?(file)
+      @config['algolia']['excluded_files'].include?(file.name)
+    end
+
     def check_credentials(api_key, application_id, index_name)
       unless api_key
         Jekyll.logger.error 'Algolia Error: No API key defined'
@@ -84,43 +88,34 @@ class AlgoliaSearchJekyllPush < Jekyll::Command
     end
 
     def configure_index(index)
-      index.set_settings(
+      default_settings = {
         attributeForDistinct: 'parent_id',
         attributesForFaceting: %w(tags type),
-        attributesToHighlight: %w(title content),
-        attributesToIndex: %w(title h1 h2 h3 h4 h5 h6 content tags),
-        attributesToRetrieve: %w(title posted_at content url css_selector),
+        attributesToIndex: %w(
+          title h1 h2 h3 h4 h5 h6
+          unordered(text)
+          unordered(tags)
+        ),
+        attributesToRetrieve: %w(
+          title h1 h2 h3 h4 h5 h6
+          posted_at
+          content
+          text
+          url
+          css_selector
+        ),
         customRanking: ['desc(posted_at)', 'desc(title_weight)'],
         distinct: true,
         highlightPreTag: '<span class="algolia__result-highlight">',
         highlightPostTag: '</span>'
-      )
-    end
-
-    def push(items)
-      api_key = AlgoliaSearchJekyll.api_key
-      application_id = @config['algolia']['application_id']
-      index_name = @config['algolia']['index_name']
-      check_credentials(api_key, application_id, index_name)
-
-      Algolia.init(application_id: application_id, api_key: api_key)
-      index = Algolia::Index.new(index_name)
-      configure_index(index)
-      index.clear_index
-
-      items.each_slice(1000) do |batch|
-        Jekyll.logger.info "Indexing #{batch.size} items"
-        begin
-          index.add_objects(batch)
-        rescue StandardError => error
-          Jekyll.logger.error 'Algolia Error: HTTP Error'
-          Jekyll.logger.warn error.message
-          exit 1
-        end
+      }
+      custom_settings = {}
+      @config['algolia']['settings'].each do |key, value|
+        custom_settings[key.to_sym] = value
       end
+      settings = default_settings.merge(custom_settings)
 
-      Jekyll.logger.info "Indexing of #{items.size} items " \
-                         "in #{index_name} done."
+      index.set_settings(settings)
     end
 
     def get_items_from_file(file)
@@ -130,6 +125,7 @@ class AlgoliaSearchJekyllPush < Jekyll::Command
       # We only index posts, and markdown pages
       return nil unless is_page || is_post
       return nil if is_page && !parseable?(file)
+      return nil if excluded_file?(file)
 
       html = file.content.gsub("\n", ' ')
 
@@ -165,6 +161,7 @@ class AlgoliaSearchJekyllPush < Jekyll::Command
       tags.map! { |tag| tag.to_s.gsub(',', '') }
     end
 
+    # Get the list of headings (h1, h2, etc) above the specified node
     def get_previous_hx(node, memo = { level: 7 })
       previous = node.previous_sibling
       # Stop if no previous element
@@ -189,7 +186,7 @@ class AlgoliaSearchJekyllPush < Jekyll::Command
       memo[:level] = title_level
 
       # Add to the memo and continue
-      memo[tag_name.to_sym] = previous.text
+      memo[tag_name.to_sym] = previous.content
       get_previous_hx(previous, memo)
     end
 
@@ -214,17 +211,47 @@ class AlgoliaSearchJekyllPush < Jekyll::Command
       node.css_path.gsub('html > body > ', '')
     end
 
+    # Get a list of items representing the different paragraphs
     def get_paragraphs_from_html(html, base_data)
       doc = Nokogiri::HTML(html)
-      doc.css('p').map.with_index do |p, index|
+      paragraphs = doc.css('p').map.with_index do |p, index|
+        next unless p.text.size > 0
         new_item = base_data.clone
         new_item.merge!(get_previous_hx(p))
         new_item[:objectID] = "#{new_item[:parent_id]}_#{index}"
         new_item[:css_selector] = get_css_selector(p)
-        new_item[:content] = p.to_s
+        new_item[:raw_html] = p.to_s
+        new_item[:text] = p.content
         new_item[:title_weight] = get_title_weight(p.text, new_item)
         new_item
       end
+      paragraphs.compact
+    end
+
+    def push(items)
+      api_key = AlgoliaSearchJekyll.api_key
+      application_id = @config['algolia']['application_id']
+      index_name = @config['algolia']['index_name']
+      check_credentials(api_key, application_id, index_name)
+
+      Algolia.init(application_id: application_id, api_key: api_key)
+      index = Algolia::Index.new(index_name)
+      configure_index(index)
+      index.clear_index
+
+      items.each_slice(1000) do |batch|
+        Jekyll.logger.info "Indexing #{batch.size} items"
+        begin
+          index.add_objects(batch)
+        rescue StandardError => error
+          Jekyll.logger.error 'Algolia Error: HTTP Error'
+          Jekyll.logger.warn error.message
+          exit 1
+        end
+      end
+
+      Jekyll.logger.info "Indexing of #{items.size} items " \
+                         "in #{index_name} done."
     end
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: algoliasearch-jekyll
 version: !ruby/object:Gem::Version
-  version: 0.1.2
+  version: 0.1.3
 platform: ruby
 authors:
 - Tim Carry
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-06-22 00:00:00.000000000 Z
+date: 2015-06-30 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: nokogiri