lib/spidr/headers.rb in spidr-0.3.2 vs lib/spidr/headers.rb in spidr-0.4.0

@@ -114,10 +114,71 @@
     def content_types
       (headers['content-type'] || [])
     end
 
     #
+    # The charset included in the Content-Type.
+    #
+    # @return [String, nil]
+    #   The charset of the content.
+    #
+    # @since 0.4.0
+    #
+    def content_charset
+      content_types.each do |value|
+        if value.include?(';')
+          value.split(';').each do |param|
+            param.strip!
+
+            if param.start_with?('charset=')
+              return param.split('=',2).last
+            end
+          end
+        end
+      end
+
+      return nil
+    end
+
+    #
+    # Determines if any of the content-types of the page include a given
+    # type.
+    #
+    # @param [String] type
+    #   The content-type to test for.
+    #
+    # @return [Boolean]
+    #   Specifies whether the page includes the given content-type.
+    #
+    # @example Match the Content-Type
+    #   page.is_content_type?('application/json')
+    #
+    # @example Match the sub-type of the Content-Type
+    #   page.is_content_type?('json')
+    #
+    # @since 0.4.0
+    #
+    def is_content_type?(type)
+      if type.include?('/')
+        # otherwise only match the first param
+        content_types.any? do |value|
+          value = value.split(';',2).first
+
+          value == type
+        end
+      else
+        # otherwise only match the sub-type
+        content_types.any? do |value|
+          value = value.split(';',2).first
+          value = value.split('/',2).last
+
+          value == type
+        end
+      end
+    end
+
+    #
     # Determines if the page is plain-text.
     #
     # @return [Boolean]
     #   Specifies whether the page is plain-text.
     #
@@ -289,37 +350,21 @@
     # @since 0.2.2
     #
     def cookie_params
       params = {}
 
-      cookies.each do |cookie|
-        cookie.split('; ').each do |key_value|
-          key, value = key_value.split('=',2)
+      cookies.each do |value|
+        value.split(';').each do |param|
+          param.strip!
 
-          unless RESERVED_COOKIE_NAMES.include?(key)
-            params[key] = (value || '')
+          name, value = param.split('=',2)
+
+          unless RESERVED_COOKIE_NAMES.include?(name)
+            params[name] = (value || '')
           end
         end
       end
 
       return params
-    end
-
-    protected
-
-    #
-    # Determines if any of the content-types of the page include a given
-    # type.
-    #
-    # @param [String] type
-    #   The content-type to test for.
-    #
-    # @return [Boolean]
-    #   Specifies whether the page includes the given content-type.
-    #
-    # @since 0.2.4
-    #
-    def is_content_type?(type)
-      content_types.any? { |content| content.include?(type) }
     end
   end
 end