Extract a common superclass out of the Python EntriesFilter classes

lib/docs/filters/python/entries_common.rb

@@ -0,0 +1,78 @@
+module Docs
+  class Python
+    class CommonEntriesFilter < Docs::EntriesFilter
+      def get_name
+        name = at_css('h1').content
+        name.remove! %r{\A[\d\.]+ } # remove list number
+        name.remove! "\u{00B6}" # remove pilcrow sign
+        name.remove! %r{ [\u{2013}\u{2014}].+\z} # remove text after em/en dash
+        name.remove! 'Built-in'
+        name.strip!
+        name
+      end
+
+      def get_type
+        return 'Logging' if slug.start_with? 'library/logging'
+
+        type = at_css('.related a[accesskey="U"]').content
+
+        original_type = type
+        type = parse_type type
+
+        if type == original_type
+          if type == 'The Python Standard Library'
+            type = at_css('h1').content
+          elsif type.include?('I/O') || %w(select selectors).include?(name)
+            type = 'Input/ouput'
+          end
+        end
+
+        type.remove! %r{\A\d+\.\s+} # remove list number
+        type.remove! "\u{00b6}" # remove paragraph character
+        type.sub! ' and ', ' & '
+        [' Services', ' Modules', ' Specific', 'Python '].each { |str| type.remove!(str) }
+
+        self.class.const_get(:REPLACE_TYPES)[type] || type
+      end
+
+      def parse_type
+        raise Error.new('Python::CommonEntriesFilter is an abstract filter. Subclass it before using it.')
+      end
+
+      def include_default_entry?
+        !at_css('.body > .section:only-child > .toctree-wrapper:last-child') && type != 'Superseded'
+      end
+
+      def additional_entries
+        return [] if root_page? || !include_default_entry? || name == 'errno'
+        clean_id_attributes
+        entries = []
+
+        css('.class > dt[id]', '.exception > dt[id]', '.attribute > dt[id]').each do |node|
+          entries << [node['id'], node['id']]
+        end
+
+        css('.data > dt[id]').each do |node|
+          if node['id'].split('.').last.upcase! # skip constants
+            entries << [node['id'], node['id']]
+          end
+        end
+
+        css('.function > dt[id]', '.method > dt[id]', '.staticmethod > dt[id]', '.classmethod > dt[id]').each do |node|
+          entries << [node['id'] + '()', node['id']]
+        end
+
+        entries
+      end
+
+      def clean_id_attributes
+        css('.section > .target[id]').each do |node|
+          if dt = node.at_css('+ dl > dt')
+            dt['id'] ||= node['id'].remove(/\w+\-/)
+          end
+          node.remove
+        end
+      end
+    end
+  end
+end

lib/docs/filters/python/entries_v2.rb

@@ -1,6 +1,6 @@
 module Docs
   class Python
-    class EntriesV2Filter < Docs::EntriesFilter
+    class EntriesV2Filter < CommonEntriesFilter
       REPLACE_TYPES = {
         'compiler package'                        => 'Compiler',
         'Cryptographic'                           => 'Cryptography',
@@ -15,72 +15,18 @@ module Docs
         'Program Frameworks'                      => 'Frameworks',
         'Structured Markup Processing Tools'      => 'Structured Markup' }
 
-      def get_name
+      def parse_type(type)
-        name = at_css('h1').content
+        if type.start_with? '18'
-        name.remove! %r{\A[\d\.]+ } # remove list number
-        name.remove! "\u{00B6}" # remove pilcrow sign
-        name.remove! %r{ [\u{2013}\u{2014}].+\z} # remove text after em/en dash
-        name.remove! 'Built-in'
-        name.strip!
-        name
-      end
-
-      def get_type
-        return 'Logging' if slug.start_with? 'library/logging'
-
-        type = at_css('.related a[accesskey="U"]').content
-
-        if type == 'The Python Standard Library'
-          type = at_css('h1').content
-        elsif type.include?('I/O') || %w(select selectors).include?(name)
-          type = 'Input/ouput'
-        elsif type.start_with? '18'
           type = 'Internet Data Handling'
         elsif type.include? 'Mac'
           type = 'Mac OS'
         end
 
-        type.remove! %r{\A\d+\.\s+} # remove list number
+        type
-        type.remove! "\u{00b6}" # remove paragraph character
-        type.sub! ' and ', ' & '
-        [' Services', ' Modules', ' Specific', 'Python '].each { |str| type.remove!(str) }
-
-        REPLACE_TYPES[type] || type
       end
 
       def include_default_entry?
-        !at_css('.body > .section:only-child > .toctree-wrapper:last-child') && !type.in?(%w(Superseded SunOS))
+        super && type != 'SunOS'
-      end
-
-      def additional_entries
-        return [] if root_page? || !include_default_entry? || name == 'errno'
-        clean_id_attributes
-        entries = []
-
-        css('.class > dt[id]', '.exception > dt[id]', '.attribute > dt[id]').each do |node|
-          entries << [node['id'], node['id']]
-        end
-
-        css('.data > dt[id]').each do |node|
-          if node['id'].split('.').last.upcase! # skip constants
-            entries << [node['id'], node['id']]
-          end
-        end
-
-        css('.function > dt[id]', '.method > dt[id]', '.staticmethod > dt[id]', '.classmethod > dt[id]').each do |node|
-          entries << [node['id'] + '()', node['id']]
-        end
-
-        entries
-      end
-
-      def clean_id_attributes
-        css('.section > .target[id]').each do |node|
-          if dt = node.at_css('+ dl > dt')
-            dt['id'] ||= node['id'].remove(/\w+\-/)
-          end
-          node.remove
-        end
       end
     end
   end

lib/docs/filters/python/entries_v3.rb

@@ -1,6 +1,6 @@
 module Docs
   class Python
-    class EntriesV3Filter < Docs::EntriesFilter
+    class EntriesV3Filter < CommonEntriesFilter
       REPLACE_TYPES = {
         'Cryptographic'                           => 'Cryptography',
         'Custom Interpreters'                     => 'Interpreters',
@@ -13,70 +13,12 @@ module Docs
         'Program Frameworks'                      => 'Frameworks',
         'Structured Markup Processing Tools'      => 'Structured Markup' }
 
-      def get_name
+      def parse_type(type)
-        name = at_css('h1').content
+        if type.start_with? '19'
-        name.remove! %r{\A[\d\.]+ } # remove list number
-        name.remove! "\u{00B6}" # remove pilcrow sign
-        name.remove! %r{ [\u{2013}\u{2014}].+\z} # remove text after em/en dash
-        name.remove! 'Built-in'
-        name.strip!
-        name
-      end
-
-      def get_type
-        return 'Logging' if slug.start_with? 'library/logging'
-
-        type = at_css('.related a[accesskey="U"]').content
-
-        if type == 'The Python Standard Library'
-          type = at_css('h1').content
-        elsif type.include?('I/O') || %w(select selectors).include?(name)
-          type = 'Input/ouput'
-        elsif type.start_with? '19'
           type = 'Internet Data Handling'
         end
 
-        type.remove! %r{\A\d+\.\s+} # remove list number
+        type
-        type.remove! "\u{00b6}" # remove paragraph character
-        type.sub! ' and ', ' & '
-        [' Services', ' Modules', ' Specific', 'Python '].each { |str| type.remove!(str) }
-
-        REPLACE_TYPES[type] || type
-      end
-
-      def include_default_entry?
-        !at_css('.body > .section:only-child > .toctree-wrapper:last-child') && !type.in?(%w(Superseded))
-      end
-
-      def additional_entries
-        return [] if root_page? || !include_default_entry? || name == 'errno'
-        clean_id_attributes
-        entries = []
-
-        css('.class > dt[id]', '.exception > dt[id]', '.attribute > dt[id]').each do |node|
-          entries << [node['id'], node['id']]
-        end
-
-        css('.data > dt[id]').each do |node|
-          if node['id'].split('.').last.upcase! # skip constants
-            entries << [node['id'], node['id']]
-          end
-        end
-
-        css('.function > dt[id]', '.method > dt[id]', '.staticmethod > dt[id]', '.classmethod > dt[id]').each do |node|
-          entries << [node['id'] + '()', node['id']]
-        end
-
-        entries
-      end
-
-      def clean_id_attributes
-        css('.section > .target[id]').each do |node|
-          if dt = node.at_css('+ dl > dt')
-            dt['id'] ||= node['id'].remove(/\w+\-/)
-          end
-          node.remove
-        end
       end
     end
   end