article_json 0.1.0

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

@@ -0,0 +1,89 @@
+module ArticleJSON
+  module Import
+    module GoogleDoc
+      module HTML
+        class TextParser
+          # @param [Nokogiri::HTML::Node] node
+          # @param [ArticleJSON::Import::GoogleDoc::HTML::CSSAnalyzer] css_analyzer
+          def initialize(node:, css_analyzer:)
+            @node = node
+            @css_analyzer = css_analyzer
+          end
+
+          # The content of the text node, w/o any markup
+          # @return [String]
+          def content
+            @node.inner_text
+          end
+
+          # Check if the text node is styled as bold
+          # @return [Boolean]
+          def bold?
+            @node.name == 'span' &&
+              @node.has_attribute?('class') &&
+              @css_analyzer.bold?(@node.attribute('class').value)
+          end
+
+          # Check if the text node is styled as italic
+          # @return [Boolean]
+          def italic?
+            @node.name == 'span' &&
+              @node.has_attribute?('class') &&
+              @css_analyzer.italic?(@node.attribute('class').value)
+          end
+
+          # A possible link target for the text, otherwise `nil`
+          # Google redirects (basically all links in a google doc html export)
+          # are stripped.
+          # @return [String]
+          def href
+            if @node.name == 'span' &&
+                @node.first_element_child&.name == 'a' &&
+                @node.first_element_child&.has_attribute?('href')
+              strip_google_redirect(
+                @node.first_element_child.attribute('href').value
+              )
+            end
+          end
+
+          # @return [ArticleJSON::Elements::Text]
+          def element
+            ArticleJSON::Elements::Text.new(
+              content: content,
+              bold: bold?,
+              italic: italic?,
+              href: href
+            )
+          end
+
+          class << self
+            # Extract multiple text nodes from a wrapping node
+            # The wrapping node is usually a paragraph or caption
+            # @param [Nokogiri::HTML::Node] node
+            # @param [ArticleJSON::Import::GoogleDoc::HTML::CSSAnalyzer] css_analyzer
+            # @return [Array[ArticleJSON::Elements::Text]]
+            def extract(node:, css_analyzer:)
+              node.children.map do |child_node|
+                next if NodeAnalyzer.new(child_node).empty?
+                new(node: child_node, css_analyzer: css_analyzer).element
+              end.compact
+            end
+          end
+
+          private
+
+          # @param [String] url
+          # @return [String]
+          def strip_google_redirect(url)
+            uri = URI(url)
+            if uri.host && uri.host.match(/google\.com/) && uri.path == '/url'
+              params = CGI.parse(uri.query)
+              return params['q'].first if params['q'] && params['q'].any?
+            end
+            url
+          end
+        end
+      end
+    end
+  end
+end

data/lib/article_json/utils/o_embed_resolver/base.rb

@@ -0,0 +1,63 @@
+module ArticleJSON
+  module Utils
+    module OEmbedResolver
+      class Base
+        # @param [ArticleJSON::Elements::Embed] embed_element
+        def initialize(embed_element)
+          @element = embed_element
+        end
+
+        # @return [Hash]
+        def oembed_data
+          resolver = self.class == Base ? self.class.build(@element) : self
+          resolver.parsed_api_response
+        end
+
+        protected
+
+        # @return [Hash]
+        def parsed_api_response
+          @api_response ||= begin
+            uri = URI.parse(oembed_url)
+            http = Net::HTTP.new(uri.host, uri.port)
+            http.use_ssl = (uri.scheme == 'https')
+            response = http.request(Net::HTTP::Get.new(uri, http_headers))
+            JSON.parse(response.body, symbolize_names: true)
+          end
+        end
+
+        # @return [Hash]
+        def http_headers
+          headers = { 'Content-Type' => 'application/json' }
+          unless ArticleJSON.configuration.oembed_user_agent.nil?
+            headers['User-Agent'] = ArticleJSON.configuration.oembed_user_agent
+          end
+          headers
+        end
+
+        class << self
+          # Instantiate the correct sub class for a given element
+          # @param [ArticleJSON::Elements::Embed] embed_element
+          # @return [ArticleJSON::Utils::OEmbedResolver::Base]
+          def build(embed_element)
+            resolver = resolver_by_embed_type(embed_element.embed_type)
+            resolver.new(embed_element) unless resolver.nil?
+          end
+
+          # Lookup the correct sub class for a given element type
+          # @param [Symbol] :type
+          # @return [ArticleJSON::Utils::OEmbedResolver::Base]
+          def resolver_by_embed_type(type)
+            {
+              facebook_video: FacebookVideo,
+              slideshare: Slideshare,
+              tweet: Tweet,
+              vimeo_video: VimeoVideo,
+              youtube_video: YoutubeVideo,
+            }[type.to_sym]
+          end
+        end
+      end
+    end
+  end
+end

data/lib/article_json/utils/o_embed_resolver/facebook_video.rb

@@ -0,0 +1,21 @@
+module ArticleJSON
+  module Utils
+    module OEmbedResolver
+      class FacebookVideo < Base
+        # The URL for the oembed API call
+        # @return [String]
+        def oembed_url
+          "https://www.facebook.com/plugins/video/oembed.json?url=#{video_url}"
+        end
+
+        private
+
+        # The video URL of the element
+        # @return [String]
+        def video_url
+          "facebook.com/facebook/videos/#{@element.embed_id}"
+        end
+      end
+    end
+  end
+end

data/lib/article_json/utils/o_embed_resolver/slideshare.rb

@@ -0,0 +1,22 @@
+module ArticleJSON
+  module Utils
+    module OEmbedResolver
+      class Slideshare < Base
+        # The URL for the oembed API call
+        # @return [String]
+        def oembed_url
+          "https://www.slideshare.net/api/oembed/2?format=json&url=#{slide_url}"
+        end
+
+        private
+
+        # The URL of the slideshow
+        # @return [String]
+        def slide_url
+          handle, slug = @element.embed_id.split('/', 2)
+          "www.slideshare.net/#{handle}/#{slug}"
+        end
+      end
+    end
+  end
+end

data/lib/article_json/utils/o_embed_resolver/tweet.rb

@@ -0,0 +1,23 @@
+module ArticleJSON
+  module Utils
+    module OEmbedResolver
+      class Tweet < Base
+        # The URL for the oembed API call
+        # @return [String]
+        def oembed_url
+          'https://api.twitter.com/1/statuses/oembed.json?align=center' \
+            "&url=#{tweet_url}"
+        end
+
+        private
+
+        # The URL of the tweet
+        # @return [String]
+        def tweet_url
+          handle, tweet_id = @element.embed_id.split('/', 2)
+          "https://twitter.com/#{handle}/status/#{tweet_id}"
+        end
+      end
+    end
+  end
+end

data/lib/article_json/utils/o_embed_resolver/vimeo_video.rb

@@ -0,0 +1,21 @@
+module ArticleJSON
+  module Utils
+    module OEmbedResolver
+      class VimeoVideo < Base
+        # The URL for the oembed API call
+        # @return [String]
+        def oembed_url
+          "https://vimeo.com/api/oembed.json?url=#{video_url}"
+        end
+
+        private
+
+        # The video URL of the element
+        # @return [String]
+        def video_url
+          "https://vimeo.com/#{@element.embed_id}"
+        end
+      end
+    end
+  end
+end

data/lib/article_json/utils/o_embed_resolver/youtube_video.rb

@@ -0,0 +1,21 @@
+module ArticleJSON
+  module Utils
+    module OEmbedResolver
+      class YoutubeVideo < Base
+        # The URL for the oembed API call
+        # @return [String]
+        def oembed_url
+          "http://www.youtube.com/oembed?format=json&url=#{video_url}"
+        end
+
+        private
+
+        # The video URL of the element
+        # @return [String]
+        def video_url
+          "youtube.com/watch?v=#{@element.embed_id}"
+        end
+      end
+    end
+  end
+end

data/lib/article_json/utils.rb

@@ -0,0 +1,11 @@
+module ArticleJSON
+  module Utils
+  end
+end
+
+require_relative 'utils/o_embed_resolver/base'
+require_relative 'utils/o_embed_resolver/facebook_video'
+require_relative 'utils/o_embed_resolver/slideshare'
+require_relative 'utils/o_embed_resolver/tweet'
+require_relative 'utils/o_embed_resolver/vimeo_video'
+require_relative 'utils/o_embed_resolver/youtube_video'

data/lib/article_json/version.rb

@@ -0,0 +1,3 @@
+module ArticleJSON
+  VERSION = '0.1.0'
+end

data/lib/article_json.rb

@@ -0,0 +1,55 @@
+require 'uri'
+require 'cgi'
+require 'json'
+require 'net/http'
+
+require 'nokogiri'
+require 'css_parser'
+
+require_relative 'article_json/version'
+require_relative 'article_json/configuration'
+require_relative 'article_json/utils'
+
+require_relative 'article_json/elements/base'
+require_relative 'article_json/elements/text'
+require_relative 'article_json/elements/heading'
+require_relative 'article_json/elements/paragraph'
+require_relative 'article_json/elements/list'
+require_relative 'article_json/elements/image'
+require_relative 'article_json/elements/text_box'
+require_relative 'article_json/elements/quote'
+require_relative 'article_json/elements/embed'
+
+require_relative 'article_json/import/google_doc/html/shared/caption'
+require_relative 'article_json/import/google_doc/html/shared/float'
+require_relative 'article_json/import/google_doc/html/css_analyzer'
+require_relative 'article_json/import/google_doc/html/node_analyzer'
+require_relative 'article_json/import/google_doc/html/text_parser'
+require_relative 'article_json/import/google_doc/html/heading_parser'
+require_relative 'article_json/import/google_doc/html/paragraph_parser'
+require_relative 'article_json/import/google_doc/html/list_parser'
+require_relative 'article_json/import/google_doc/html/image_parser'
+require_relative 'article_json/import/google_doc/html/text_box_parser'
+require_relative 'article_json/import/google_doc/html/quote_parser'
+require_relative 'article_json/import/google_doc/html/embedded_parser'
+require_relative 'article_json/import/google_doc/html/embedded_facebook_video_parser'
+require_relative 'article_json/import/google_doc/html/embedded_vimeo_video_parser'
+require_relative 'article_json/import/google_doc/html/embedded_youtube_video_parser'
+require_relative 'article_json/import/google_doc/html/embedded_slideshare_parser'
+require_relative 'article_json/import/google_doc/html/embedded_tweet_parser'
+require_relative 'article_json/import/google_doc/html/parser'
+
+require_relative 'article_json/export/html/elements/shared/caption'
+require_relative 'article_json/export/html/elements/shared/float'
+require_relative 'article_json/export/html/elements/base'
+require_relative 'article_json/export/html/elements/text'
+require_relative 'article_json/export/html/elements/heading'
+require_relative 'article_json/export/html/elements/paragraph'
+require_relative 'article_json/export/html/elements/list'
+require_relative 'article_json/export/html/elements/image'
+require_relative 'article_json/export/html/elements/text_box'
+require_relative 'article_json/export/html/elements/quote'
+require_relative 'article_json/export/html/elements/embed'
+require_relative 'article_json/export/html/exporter'
+
+require_relative 'article_json/article'

metadata

@@ -0,0 +1,189 @@
+--- !ruby/object:Gem::Specification
+name: article_json
+version: !ruby/object:Gem::Version
+  version: 0.1.0
+platform: ruby
+authors:
+- "@dsager"
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2017-09-20 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: nokogiri
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.8'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.8'
+- !ruby/object:Gem::Dependency
+  name: css_parser
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.5'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.5'
+- !ruby/object:Gem::Dependency
+  name: bundler
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.15'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.15'
+- !ruby/object:Gem::Dependency
+  name: rspec
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.6'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.6'
+- !ruby/object:Gem::Dependency
+  name: webmock
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.0'
+- !ruby/object:Gem::Dependency
+  name: coveralls
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.8'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.8'
+description: |
+  `article_json` is a format definition for news articles and a ruby gem that
+  offers conversions from and to different formats:
+  - Parser for Google Doc HTML exports
+  - Converter to simple HTML format
+  - Converter to AMP format
+email: info@devex.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- CHANGELOG.md
+- LICENSE
+- README.md
+- bin/article_json_export_google_doc.rb
+- bin/article_json_export_html.rb
+- bin/article_json_parse_google_doc.rb
+- bin/update_reference_document.sh
+- lib/article_json.rb
+- lib/article_json/article.rb
+- lib/article_json/configuration.rb
+- lib/article_json/elements/base.rb
+- lib/article_json/elements/embed.rb
+- lib/article_json/elements/heading.rb
+- lib/article_json/elements/image.rb
+- lib/article_json/elements/list.rb
+- lib/article_json/elements/paragraph.rb
+- lib/article_json/elements/quote.rb
+- lib/article_json/elements/text.rb
+- lib/article_json/elements/text_box.rb
+- lib/article_json/export/html/elements/base.rb
+- lib/article_json/export/html/elements/embed.rb
+- lib/article_json/export/html/elements/heading.rb
+- lib/article_json/export/html/elements/image.rb
+- lib/article_json/export/html/elements/list.rb
+- lib/article_json/export/html/elements/paragraph.rb
+- lib/article_json/export/html/elements/quote.rb
+- lib/article_json/export/html/elements/shared/caption.rb
+- lib/article_json/export/html/elements/shared/float.rb
+- lib/article_json/export/html/elements/text.rb
+- lib/article_json/export/html/elements/text_box.rb
+- lib/article_json/export/html/exporter.rb
+- lib/article_json/import/google_doc/html/css_analyzer.rb
+- lib/article_json/import/google_doc/html/embedded_facebook_video_parser.rb
+- lib/article_json/import/google_doc/html/embedded_parser.rb
+- lib/article_json/import/google_doc/html/embedded_slideshare_parser.rb
+- lib/article_json/import/google_doc/html/embedded_tweet_parser.rb
+- lib/article_json/import/google_doc/html/embedded_vimeo_video_parser.rb
+- lib/article_json/import/google_doc/html/embedded_youtube_video_parser.rb
+- lib/article_json/import/google_doc/html/heading_parser.rb
+- lib/article_json/import/google_doc/html/image_parser.rb
+- lib/article_json/import/google_doc/html/list_parser.rb
+- lib/article_json/import/google_doc/html/node_analyzer.rb
+- lib/article_json/import/google_doc/html/paragraph_parser.rb
+- lib/article_json/import/google_doc/html/parser.rb
+- lib/article_json/import/google_doc/html/quote_parser.rb
+- lib/article_json/import/google_doc/html/shared/caption.rb
+- lib/article_json/import/google_doc/html/shared/float.rb
+- lib/article_json/import/google_doc/html/text_box_parser.rb
+- lib/article_json/import/google_doc/html/text_parser.rb
+- lib/article_json/utils.rb
+- lib/article_json/utils/o_embed_resolver/base.rb
+- lib/article_json/utils/o_embed_resolver/facebook_video.rb
+- lib/article_json/utils/o_embed_resolver/slideshare.rb
+- lib/article_json/utils/o_embed_resolver/tweet.rb
+- lib/article_json/utils/o_embed_resolver/vimeo_video.rb
+- lib/article_json/utils/o_embed_resolver/youtube_video.rb
+- lib/article_json/version.rb
+homepage: https://github.com/Devex/article_json
+licenses:
+- MIT
+metadata: {}
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '2.3'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 2.5.1
+signing_key: 
+specification_version: 4
+summary: JSON Format for News Articles & Ruby Gem
+test_files: []