auto_source.rb in html2rss-0.14.0

- old
+ new

@@ -14,28 +14,32 @@
     class UnsupportedUrlScheme < Html2rss::Error; end
     class NoArticlesFound < Html2rss::Error; end
 
     SUPPORTED_URL_SCHEMES = %w[http https].to_set.freeze
 
-    def initialize(url)
-      unless url.is_a?(String) || url.is_a?(Addressable::URI)
-        raise ArgumentError,
-              'URL must be a String or Addressable::URI'
-      end
+    ##
+    # @param url [Addressable::URI] The URL to extract articles from.
+    # @param body [String] The body of the response.
+    # @param headers [Hash] The headers of the response.
+    def initialize(url, body:, headers: {})
+      raise ArgumentError, 'URL must be a Addressable::URI' unless url.is_a?(Addressable::URI)
+      raise ArgumentError, 'URL must be absolute' unless url.absolute?
+      raise UnsupportedUrlScheme, "#{url.scheme} not supported" unless SUPPORTED_URL_SCHEMES.include?(url.scheme)
 
-      @url = Addressable::URI.parse(url)
-
-      raise ArgumentError, 'URL must be absolute' unless @url.absolute?
-      raise UnsupportedUrlScheme, "#{@url.scheme} not supported" unless SUPPORTED_URL_SCHEMES.include?(@url.scheme)
+      @url = url
+      @body = body
+      @headers = headers
     end
 
     def build
       raise NoArticlesFound if articles.empty?
 
       Reducer.call(articles, url:)
       Cleanup.call(articles, url:, keep_different_domain: true)
 
+      channel.articles = articles
+
       Html2rss::AutoSource::RssBuilder.new(
         channel:,
         articles:
       ).call
     end
@@ -55,23 +59,22 @@
         articles_in_thread
       end
     end
 
     def channel
-      Channel.new(parsed_body, response:, url:, articles:)
+      @channel ||= Channel.new(parsed_body, headers: @headers, url:)
     end
 
     private
 
     attr_reader :url
 
-    def response
-      @response ||= Html2rss::Utils.request_url(url)
-    end
-
-    # Parses the HTML body of the response using Nokogiri.
     # @return [Nokogiri::HTML::Document]
     def parsed_body
-      @parsed_body ||= Nokogiri.HTML(response.body).freeze
+      @parsed_body ||= Nokogiri.HTML(@body)
+                               .tap do |doc|
+        # Remove comments from the document
+        doc.xpath('//comment()').each(&:remove)
+      end.freeze
     end
   end
 end