article_json 0.3.5 → 0.3.6

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: f34cc736d928d12ee64d0181a15d9b69b2bdfc6c89e76c35ffd79cb6a8acb8d5
-  data.tar.gz: 74c2c855413da0691866fe54d1927b4342bb5409a826807821866613203f3db9
+  metadata.gz: b7f9acdba2ed14993cb2ce87317b205c65992e79ed5662d79d6d3fb357908ff1
+  data.tar.gz: b7638f1a409aa31a52a911001684eeda751f254ecb357b4907710f24b97cbf95
 SHA512:
-  metadata.gz: ec2b0a1718c9f57dd278ce4e11bcca41a84353b3594be06599204b838701199ecaf5e21bb36921ea68c98c7fa7083d2cd8f939fb7253b07305aa4258bbf47c25
-  data.tar.gz: 381aad7bba231456aa45bf4f512bd2fa788dffa875b1eccf5a77d859cff28d5b061eb3b7a08c0e51b0915a57c9695c8113d07966b313be0d9ca00ad8759f2469
+  metadata.gz: 297a7a8f4445d1525803cb3c57cad514e43cedc2c3c354d1c98d5fce603c78908dd7f3173f3879ab2f00cbce6c06b2c71c5a1261a046613a25f59fbfeb8cd6c4
+  data.tar.gz: af71578a715e7245563d3e204e9c767c45adbe56e5697ab320b7c736ad18259bb7111a90453a63d1ac8bd75b7139616823a35d6a039f16777435cb17aa848696

data/CHANGELOG.md

@@ -1,4 +1,7 @@
 # Changelog
+## 0.3.6 - 2019/8/6
+- **Improvement** Added tags support in text_box element.
+
 ## 0.3.5 - 2018/12/12
 - **Improvements** to import and export image links from Google Docs
 - Import image `href`` from caption text using a custom tag

data/lib/article_json/elements/text_box.rb

@@ -1,14 +1,16 @@
 module ArticleJSON
   module Elements
     class TextBox < Base
-      attr_reader :content, :float
+      attr_reader :content, :float, :tags
 
       # @param [Array[Paragraph|Heading|List]] content
       # @param [Symbol] float
-      def initialize(content:, float: nil)
+      # @param [Array] tags
+      def initialize(content:, float: nil, tags: [])
         @type = :text_box
         @content = content
         @float = float
+        @tags = tags
       end
 
       # Hash representation of this text box element
@@ -18,6 +20,7 @@ module ArticleJSON
           type: type,
           float: float,
           content: content.map(&:to_h),
+          tags: tags,
         }
       end
 
@@ -27,7 +30,8 @@ module ArticleJSON
         def parse_hash(hash)
           new(
             content: parse_hash_list(hash[:content]),
-            float: hash[:float]&.to_sym
+            float: hash[:float]&.to_sym,
+            tags: hash[:tags]
           )
         end
       end

data/lib/article_json/export/common/html/elements/image.rb

@@ -19,7 +19,7 @@ module ArticleJSON
             # @return [Nokogiri::XML::NodeSet]
             def figure_node
               create_element(:figure, node_opts) do |figure|
-                node =  @element.href.present? ? href_node : image_node
+                node =  @element&.href ? href_node : image_node
                 figure.add_child(node)
                 if @element.caption&.any?
                   figure.add_child(caption_node(:figcaption))

data/lib/article_json/import/google_doc/html/node_analyzer.rb

@@ -17,6 +17,14 @@ module ArticleJSON
             node.inner_text.strip.downcase == text.strip.downcase
           end
 
+          # Check if the node text begins with a certain text
+          # @param [String]
+          # @return [Boolean]
+          def begins_with?(text)
+            first_word = node.inner_text.strip.downcase.split(' ').first
+            first_word == text.strip.downcase
+          end
+
           # Check if the node is empty, i.e. not containing any text
           # Given that images are the only nodes without text, we have to make
           # sure that it's not an image.
@@ -65,7 +73,7 @@ module ArticleJSON
           # @return [Boolean]
           def text_box?
             return @is_text_box if defined? @is_text_box
-            @is_text_box = has_text?('textbox:') || has_text?('highlight:')
+            @is_text_box = begins_with?('textbox:') || begins_with?('highlight:')
           end
 
           # Check if the node starts a quote

data/lib/article_json/import/google_doc/html/parser.rb

@@ -83,7 +83,11 @@ module ArticleJSON
           # @return [ArticleJSON::Elements::TextBox]
           def parse_text_box
             TextBoxParser
-              .new(nodes: nodes_until_hr, css_analyzer: @css_analyzer)
+              .new(
+                type_node: @current_node.node,
+                nodes: nodes_until_hr,
+                css_analyzer: @css_analyzer
+              )
               .element
           end
 

data/lib/article_json/import/google_doc/html/text_box_parser.rb

@@ -5,14 +5,18 @@ module ArticleJSON
         class TextBoxParser
           include Shared::Float
 
+          # @param [Nokogiri::HTML::Node] type_node Document node that states
+          #                                         that this is a textbox.
+          #                                         May contain tags, too.
           # @param [Array[Nokogiri::HTML::Node]] nodes
           # @param [ArticleJSON::Import::GoogleDoc::HTML::CSSAnalyzer] css_analyzer
-          def initialize(nodes:, css_analyzer:)
+          def initialize(type_node: ,nodes:, css_analyzer:)
             @nodes = nodes.reject { |node| NodeAnalyzer.new(node).empty? }
             @css_analyzer = css_analyzer
 
             # First node of the text box indicates floating behavior
             @float_node = @nodes.first
+            @type_node = type_node
           end
 
           # Parse the text box's nodes to get a list of sub elements
@@ -22,10 +26,23 @@ module ArticleJSON
             @nodes.map { |node| parse_sub_node(node) }.compact
           end
 
+          # Extract any potential tags, specified in brackets after the Textbox definition
+          # @return [Array[Symbol]]
+          def tags
+            match = /(.*?)[\s\u00A0]+\[(?<tags>.*)\]/
+                      .match(@type_node.inner_text)
+            return [] unless match
+            match[:tags].split(' ')
+          end
+
           # Hash representation of this text box
           # @return [ArticleJSON::Elements::TextBox]
           def element
-            ArticleJSON::Elements::TextBox.new(float: float, content: content)
+            ArticleJSON::Elements::TextBox.new(
+              float: float,
+              content: content,
+              tags: tags
+            )
           end
 
           private

data/lib/article_json/version.rb

@@ -1,3 +1,3 @@
 module ArticleJSON
-  VERSION = '0.3.5'
+  VERSION = '0.3.6'
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: article_json
 version: !ruby/object:Gem::Version
-  version: 0.3.5
+  version: 0.3.6
 platform: ruby
 authors:
 - Daniel Sager
@@ -10,7 +10,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2018-12-12 00:00:00.000000000 Z
+date: 2019-08-06 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: nokogiri
@@ -229,8 +229,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubyforge_project: 
-rubygems_version: 2.7.7
+rubygems_version: 3.0.4
 signing_key: 
 specification_version: 4
 summary: JSON Format for News Articles & Ruby Gem