onebox 2.2.10 → 2.2.15

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: f942eef9390a961ec26aa8d4301bf83eedabe577d067639ab6e0036ca20e02b9
-  data.tar.gz: b05c99bd27f025edc02475e38cc28b900bfa2bf64e1c5f9cd4be74cd5ab927a4
+  metadata.gz: 5e2169ef1cfb44f0208566f5ecff7bd7c5d2bbe2f12b58791f50aa24b776e56d
+  data.tar.gz: 9f9d38b578e46e6ce8c8c007577ff02522eab24f3dcc42f9eabd5b838a668e72
 SHA512:
-  metadata.gz: '081ccc4c533884804d094d40b71175409077002595b10b9c3fd56d8f6c11bcacdc99502fdd78f082adb1e59d83090182550436579ec8bc2361714bb659f7a693'
-  data.tar.gz: d3dfcb8f455e7d9189415cfdf9e00f7f1d938b8ce41d8d307ff2ff4bbb6fcf09449a4a881b0bfec34f8b275556691601f2e6d070351ea2b51d5626e5a561b2fd
+  metadata.gz: e293af57162b61ad0fa3472b7d9b35709f13504308e69eba45f40e9babff53e0258186add06ed4a89bc8305b8903dc49718c274659b5fca0712bc26d281ca7d4
+  data.tar.gz: 3bb70c552e010149bc32fd4c9dbe416cd6ae0098b4286e5894fbc12c07628b5079c0463ad78b54ab6dc773954e055c9195aea5ac180ab8bdf7dca3847db1bc96

data/.rspec

@@ -1,5 +1 @@
 --color
---format documentation
---fail-fast
---backtrace
-

data/lib/onebox/engine.rb

@@ -28,23 +28,19 @@ module Onebox
       end
     end
 
-    attr_reader :url, :uri
-    attr_reader :timeout
+    attr_reader :url, :uri, :options, :timeout
     attr :errors
 
     DEFAULT = {}
-    def options
-      @options
-    end
 
     def options=(opt)
-      return @options if opt.nil? #make sure options provided
-      opt = opt.to_h  if opt.instance_of?(OpenStruct)
+      return @options if opt.nil? # make sure options provided
+      opt = opt.to_h if opt.instance_of?(OpenStruct)
       @options.merge!(opt)
       @options
     end
 
-    def initialize(link, timeout = nil)
+    def initialize(url, timeout = nil)
       @errors = {}
       @options = DEFAULT
       class_name = self.class.name.split("::").last.to_s
@@ -52,8 +48,8 @@ module Onebox
       # Set the engine options extracted from global options.
       self.options = Onebox.options[class_name] || {}
 
-      @url = link
-      @uri = URI(link)
+      @url = url
+      @uri = URI(url)
       if always_https?
         @uri.scheme = 'https'
         @url = @uri.to_s
@@ -136,7 +132,7 @@ module Onebox
       end
 
       def always_https?
-        @https
+        defined?(@https) ? @https : false
       end
     end
   end

data/lib/onebox/engine/allowlisted_generic_onebox.rb

@@ -27,7 +27,6 @@ module Onebox
           500px.com
           8tracks.com
           abc.net.au
-          about.com
           answers.com
           arstechnica.com
           ask.com
@@ -36,11 +35,9 @@ module Onebox
           bbs.boingboing.net
           bestbuy.ca
           bestbuy.com
-          blip.tv
           bloomberg.com
           businessinsider.com
           change.org
-          clikthrough.com
           cnet.com
           cnn.com
           codepen.io
@@ -90,7 +87,6 @@ module Onebox
           meetup.com
           mixcloud.com
           mlb.com
-          myshopify.com
           myspace.com
           nba.com
           npr.org
@@ -98,16 +94,13 @@ module Onebox
           photobucket.com
           pinterest.com
           reference.com
-          revision3.com
           rottentomatoes.com
           samsung.com
-          screenr.com
           scribd.com
           slideshare.net
           sourceforge.net
           speakerdeck.com
           spotify.com
-          squidoo.com
           streamable.com
           techcrunch.com
           ted.com
@@ -124,7 +117,6 @@ module Onebox
           twitpic.com
           usatoday.com
           viddler.com
-          videojug.com
           vine.co
           walmart.com
           washingtonpost.com
@@ -275,7 +267,6 @@ module Onebox
 
       def rewrite_https(html)
         return unless html
-        uri = URI(@url)
         if AllowlistedGenericOnebox.host_matches(uri, AllowlistedGenericOnebox.rewrites)
           html = html.gsub("http://", "https://")
         end
@@ -374,15 +365,17 @@ module Onebox
         escaped_image_src = ::Onebox::Helpers.normalize_url_for_output(data[:image])
 
         <<-HTML
-            <video title='#{data[:title]}'
-                   width='#{data[:video_width]}'
-                   height='#{data[:video_height]}'
-                   style='max-width:100%'
-                   poster='#{escaped_image_src}'
-                   controls=''>
-              <source src='#{escaped_video_src}'>
-            </video>
-          HTML
+          <video
+            title='#{data[:title]}'
+            width='#{data[:video_width]}'
+            height='#{data[:video_height]}'
+            style='max-width:100%'
+            poster='#{escaped_image_src}'
+            controls=''
+          >
+            <source src='#{escaped_video_src}'>
+          </video>
+        HTML
       end
 
       def embedded_html

data/lib/onebox/engine/amazon_onebox.rb

@@ -14,13 +14,13 @@ module Onebox
       matches_regexp(/^https?:\/\/(?:www\.)?(?:smile\.)?(amazon|amzn)\.(?<tld>com|ca|de|it|es|fr|co\.jp|co\.uk|cn|in|com\.br|com\.mx|nl|pl|sa|sg|se|com\.tr|ae)\//)
 
       def url
+        @raw ||= nil
+
         # If possible, fetch the cached HTML body immediately so we can
         # try to grab the canonical URL from that document,
         # rather than guess at the best URL structure to use
-        if body_cacher&.respond_to?('cache_response_body?')
-          if body_cacher.cache_response_body?(uri.to_s) && body_cacher.cached_response_body_exists?(uri.to_s)
-            @raw ||= Onebox::Helpers.fetch_html_doc(@url, http_params, body_cacher)
-          end
+        if !@raw && has_cached_body
+          @raw = Onebox::Helpers.fetch_html_doc(@url, http_params, body_cacher)
         end
 
         if @raw
@@ -29,14 +29,15 @@ module Onebox
         end
 
         if match && match[:id]
-          return "https://www.amazon.#{tld}/dp/#{Onebox::Helpers.uri_encode(match[:id])}"
+          id = Addressable::URI.encode_component(match[:id], Addressable::URI::CharacterClasses::PATH)
+          return "https://www.amazon.#{tld}/dp/#{id}"
         end
 
         @url
       end
 
       def tld
-        @tld || @@matcher.match(@url)["tld"]
+        @tld ||= @@matcher.match(@url)["tld"]
       end
 
       def http_params
@@ -47,6 +48,12 @@ module Onebox
 
       private
 
+      def has_cached_body
+        body_cacher&.respond_to?('cache_response_body?') &&
+          body_cacher.cache_response_body?(uri.to_s) &&
+          body_cacher.cached_response_body_exists?(uri.to_s)
+      end
+
       def match
         @match ||= @url.match(/(?:d|g)p\/(?:product\/|video\/detail\/)?(?<id>[A-Z0-9]+)(?:\/|\?|$)/mi)
       end
@@ -55,9 +62,9 @@ module Onebox
         if (main_image = raw.css("#main-image")) && main_image.any?
           attributes = main_image.first.attributes
 
-          return attributes["data-a-hires"].to_s if attributes["data-a-hires"]
-
-          if attributes["data-a-dynamic-image"]
+          if attributes["data-a-hires"]
+            return attributes["data-a-hires"].to_s
+          elsif attributes["data-a-dynamic-image"]
             return ::JSON.parse(attributes["data-a-dynamic-image"].value).keys.first
           end
         end
@@ -65,9 +72,11 @@ module Onebox
         if (landing_image = raw.css("#landingImage")) && landing_image.any?
           attributes = landing_image.first.attributes
 
-          return attributes["data-old-hires"].to_s if attributes["data-old-hires"]
-
-          landing_image.first["src"].to_s
+          if attributes["data-old-hires"]
+            return attributes["data-old-hires"].to_s
+          else
+            return landing_image.first["src"].to_s
+          end
         end
 
         if (ebook_image = raw.css("#ebooksImgBlkFront")) && ebook_image.any?
@@ -89,16 +98,16 @@ module Onebox
       end
 
       def multiple_authors(authors_xpath)
-        author_list = raw.xpath(authors_xpath)
-        authors = []
-        author_list.each { |a| authors << a.inner_text.strip }
-        authors.join(", ")
+        raw
+          .xpath(authors_xpath)
+          .map { |a| a.inner_text.strip }
+          .join(", ")
       end
 
       def data
         og = ::Onebox::OpenGraph.new(raw)
 
-        if raw.at_css('#dp.book_mobile') #printed books
+        if raw.at_css('#dp.book_mobile') # printed books
           title = raw.at("h1#title")&.inner_text
           authors = raw.at_css('#byline_secondary_view_div') ? multiple_authors("//div[@id='byline_secondary_view_div']//span[@class='a-text-bold']") : raw.at("#byline")&.inner_text
           rating = raw.at("#averageCustomerReviews_feature_div .a-icon")&.inner_text || raw.at("#cmrsArcLink .a-icon")&.inner_text

data/lib/onebox/engine/asciinema_onebox.rb

@@ -7,7 +7,6 @@ module Onebox
       include StandardEmbed
 
       always_https
-
       matches_regexp(/^https?:\/\/asciinema\.org\/a\/[\p{Alnum}_\-]+$/)
 
       def to_html

data/lib/onebox/engine/audioboom_onebox.rb

@@ -13,14 +13,17 @@ module Onebox
         oembed = get_oembed
 
         <<-HTML
-          <img src="#{oembed.thumbnail_url}" style="max-width: #{oembed.width}px; max-height: #{oembed.height}px;" #{oembed.title_attr}>
+          <img
+            src="#{oembed.thumbnail_url}"
+            style="max-width: #{oembed.width}px; max-height: #{oembed.height}px;"
+            #{oembed.title_attr}
+          >
         HTML
       end
 
       def to_html
         get_oembed.html
       end
-
     end
   end
 end

data/lib/onebox/engine/bandcamp_onebox.rb

@@ -20,16 +20,16 @@ module Onebox
         escaped_src = og.video_secure_url || og.video
 
         <<-HTML
-          <iframe src="#{escaped_src}"
-                  width="#{og.video_width}"
-                  height="#{og.video_height}"
-                  scrolling="no"
-                  frameborder="0"
-                  allowfullscreen>
-          </iframe>
+          <iframe
+            src="#{escaped_src}"
+            width="#{og.video_width}"
+            height="#{og.video_height}"
+            scrolling="no"
+            frameborder="0"
+            allowfullscreen
+          ></iframe>
         HTML
       end
-
     end
   end
 end

data/lib/onebox/engine/cloudapp_onebox.rb

@@ -25,28 +25,28 @@ module Onebox
 
       def link_html(og)
         <<-HTML
-            <a href='#{og.url}' target='_blank' rel='noopener'>
-              #{og.title}
-            </a>
-          HTML
+          <a href='#{og.url}' target='_blank' rel='noopener'>
+            #{og.title}
+          </a>
+        HTML
       end
 
       def video_html(og)
         direct_src = ::Onebox::Helpers.normalize_url_for_output("#{og.get(:url)}/#{og.title}")
 
         <<-HTML
-            <video width='480' height='360' #{og.title_attr} controls loop>
-              <source src='#{direct_src}' type='video/mp4'>
-            </video>
-          HTML
+          <video width='480' height='360' #{og.title_attr} controls loop>
+            <source src='#{direct_src}' type='video/mp4'>
+          </video>
+        HTML
       end
 
       def image_html(og)
         <<-HTML
-            <a href='#{og.url}' target='_blank' class='onebox' rel='noopener'>
-              <img src='#{og.image}' #{og.title_attr} alt='CloudApp' width='480'>
-            </a>
-          HTML
+          <a href='#{og.url}' target='_blank' class='onebox' rel='noopener'>
+            <img src='#{og.image}' #{og.title_attr} alt='CloudApp' width='480'>
+          </a>
+        HTML
       end
     end
   end

data/lib/onebox/engine/coub_onebox.rb

@@ -17,7 +17,6 @@ module Onebox
       def to_html
         get_oembed.html
       end
-
     end
   end
 end

data/lib/onebox/engine/facebook_media_onebox.rb

@@ -14,13 +14,14 @@ module Onebox
         metadata = get_twitter
         if metadata.present? && metadata[:card] == "player" && metadata[:player].present?
           <<-HTML
-            <iframe src="#{metadata[:player]}"
-                    width="#{metadata[:player_width]}"
-                    height="#{metadata[:player_height]}"
-                    scrolling="no"
-                    frameborder="0"
-                    allowfullscreen>
-            </iframe>
+            <iframe
+              src="#{metadata[:player]}"
+              width="#{metadata[:player_width]}"
+              height="#{metadata[:player_height]}"
+              scrolling="no"
+              frameborder="0"
+              allowfullscreen
+            ></iframe>
           HTML
         else
           html = Onebox::Engine::AllowlistedGenericOnebox.new(@url, @timeout).to_html

data/lib/onebox/engine/five_hundred_px_onebox.rb

@@ -13,7 +13,6 @@ module Onebox
         og = get_opengraph
         "<img src='#{og.image}' width='#{og.image_width}' height='#{og.image_height}' class='onebox' #{og.title_attr}>"
       end
-
     end
   end
 end

data/lib/onebox/engine/flickr_onebox.rb

@@ -25,29 +25,28 @@ module Onebox
         album_title = "[Album] #{og.title}"
 
         <<-HTML
-            <div class='onebox flickr-album'>
-              <a href='#{escaped_url}' target='_blank' rel='noopener'>
-                <span class='outer-box' style='max-width:#{og.image_width}px'>
-                  <span class='inner-box'>
-                    <span class='album-title'>#{album_title}</span>
-                  </span>
+          <div class='onebox flickr-album'>
+            <a href='#{escaped_url}' target='_blank' rel='noopener'>
+              <span class='outer-box' style='max-width:#{og.image_width}px'>
+                <span class='inner-box'>
+                  <span class='album-title'>#{album_title}</span>
                 </span>
-                <img src='#{og.get_secure_image}' #{og.title_attr} height='#{og.image_height}' width='#{og.image_width}'>
-              </a>
-            </div>
-          HTML
+              </span>
+              <img src='#{og.secure_image_url}' #{og.title_attr} height='#{og.image_height}' width='#{og.image_width}'>
+            </a>
+          </div>
+        HTML
       end
 
       def image_html(og)
         escaped_url = ::Onebox::Helpers.normalize_url_for_output(url)
 
         <<-HTML
-            <a href='#{escaped_url}' target='_blank' rel='noopener' class="onebox">
-              <img src='#{og.get_secure_image}' #{og.title_attr} alt='Imgur' height='#{og.image_height}' width='#{og.image_width}'>
-            </a>
-          HTML
+          <a href='#{escaped_url}' target='_blank' rel='noopener' class="onebox">
+            <img src='#{og.secure_image_url}' #{og.title_attr} alt='Imgur' height='#{og.image_height}' width='#{og.image_width}'>
+          </a>
+        HTML
       end
-
     end
   end
 end

data/lib/onebox/engine/gfycat_onebox.rb

@@ -9,8 +9,8 @@ module Onebox
       matches_regexp(/^https?:\/\/gfycat\.com\//)
       always_https
 
+      # This engine should have priority over AllowlistedGenericOnebox.
       def self.priority
-        # This engine should have priority over AllowlistedGenericOnebox.
         1
       end
 
@@ -21,6 +21,7 @@ module Onebox
               <img src="https://gfycat.com/static/favicons/favicon-96x96.png" class="site-icon" width="64" height="64">
               <a href="#{data[:url]}" target="_blank" rel="nofollow ugc noopener">Gfycat.com</a>
             </header>
+
             <article class="onebox-body">
               <h4>
                 #{data[:title]} by
@@ -36,11 +37,12 @@ module Onebox
                   <img title="Sorry, your browser doesn't support HTML5 video." src="#{data[:posterUrl]}">
                 </video>
               </div>
+
               <p>
                 <span class="label1">#{data[:keywords]}</span>
               </p>
-
             </article>
+
             <div style="clear: both"></div>
           </aside>
         HTML
@@ -61,52 +63,50 @@ module Onebox
         @match ||= @url.match(/^https?:\/\/gfycat\.com\/(gifs\/detail\/)?(?<name>.+)/)
       end
 
-      def nokogiri_page
-        @nokogiri_page ||= begin
-          response = Onebox::Helpers.fetch_response(url, redirect_limit: 10) rescue nil
-          Nokogiri::HTML(response)
-        end
-      end
+      def og_data
+        return @og_data if defined?(@og_data)
 
-      def get_og_data
-        og_data = {}
+        response = Onebox::Helpers.fetch_response(url, redirect_limit: 10) rescue nil
+        page = Nokogiri::HTML(response)
+        script = page.at_css('script[type="application/ld+json"]')
 
-        if json_string = nokogiri_page.at_css('script[type="application/ld+json"]')&.text
-          og_data = Onebox::Helpers.symbolize_keys(::MultiJson.load(json_string))
+        if json_string = script&.text
+          @og_data = Onebox::Helpers.symbolize_keys(::MultiJson.load(json_string))
+        else
+          @og_data = {}
         end
-
-        og_data
       end
 
       def data
-        og_data = get_og_data
+        return @data if defined?(@data)
 
-        response = {
+        @data = {
           name: match[:name],
           title: og_data[:headline] || 'No Title',
           author: og_data[:author],
-          url: @url
+          url: @url,
         }
 
-        keywords = og_data[:keywords]&.split(',')
-        if keywords
-          response[:keywords] = keywords.map { |t| "<a href='https://gfycat.com/gifs/search/#{t}'>##{t}</a>" }.join(' ')
+        if keywords = og_data[:keywords]&.split(',')
+          @data[:keywords] = keywords
+            .map { |keyword| "<a href='https://gfycat.com/gifs/search/#{keyword}'>##{keyword}</a>" }
+            .join(' ')
         end
 
         if og_data[:video]
           content_url = ::Onebox::Helpers.normalize_url_for_output(og_data[:video][:contentUrl])
           video_url = Pathname.new(content_url)
-          response[:webmUrl] = video_url.sub_ext(".webm").to_s
-          response[:mp4Url] = video_url.sub_ext(".mp4").to_s
+          @data[:webmUrl] = video_url.sub_ext(".webm").to_s
+          @data[:mp4Url] = video_url.sub_ext(".mp4").to_s
 
           thumbnail_url = ::Onebox::Helpers.normalize_url_for_output(og_data[:video][:thumbnailUrl])
-          response[:posterUrl] = thumbnail_url
+          @data[:posterUrl] = thumbnail_url
 
-          response[:width] = og_data[:video][:width]
-          response[:height] = og_data[:video][:height]
+          @data[:width] = og_data[:video][:width]
+          @data[:height] = og_data[:video][:height]
         end
 
-        response
+        @data
       end
     end
   end