activewarehouse 0.2.0 → 0.3.0

data/README

@@ -1,6 +1,8 @@
 == ActiveWarehouse
 
-The ActiveWarehouse library provides classes and functions which help with building Data Warehouses using Rails. It can be installed either as a plugin or as a Gem.
+The ActiveWarehouse library provides classes and functions which help with
+building Data Warehouses using Rails. It can be installed either as a plugin
+or as a Gem.
 
 To install as a plugin just use:
 
@@ -11,19 +13,19 @@ To get the latest edge version.
 To install as a Gem, use:
 
   gem install activewarehouse
-  
-At this point you can use ActiveWarehouse in any application using:
 
-  require_gem 'activewarehouse'
-  require 'active_warehouse'
-  
-If you want to use the Gem and link to your Rails app then use the gemsonrails project:
+On *nix you will need to run this command as root or better yet, using sudo.
+
+Next, you will need to freeze or link the Gem to your Rails app. I prefer using
+the gemsonrails project:
 
   gem install gemsonrails
 
 And then in your Rails app:
 
   rake gems:link GEM=activewarehouse
+  
+It is possible that freezing the Gem to the Rails app may not work at all times. It is most often best to install as a plugin.
 
 == Generators
 
@@ -52,12 +54,14 @@ ActiveWarehouse comes with several generators
   script/generate dimension_view OrderDate Date
   script/generate dimension_view order_date date
   
-    Creates an OrderDateDimension class which is represented by a view on top of the DateDimension.
+    Creates an OrderDateDimension class which is represented by a view on top
+    of the DateDimension.
    
 The rules for naming are as follows:
 
 Facts:
-  Fact classes and tables follow the typical Rails rules: classes are singular and tables are pluralized. 
+  Fact classes and tables follow the typical Rails rules: classes are singular
+  and tables are pluralized. 
   Both the class and table name are suffixed by "_fact".
 Dimensions:
   Dimension classes and tables are both singular. 
@@ -68,19 +72,28 @@ Bridge:
   Bridge classes and tables are both singular.
   Both the class name and the table name are suffixed by "_bridge".
 Dimension View:
-  Dimension View classes are singular. The underlying data structure is a view on top of an existing dimension.
+  Dimension View classes are singular. The underlying data structure is a view
+  on top of an existing dimension.
   Both the class name and the view name are suffixed by "_dimension"
   
 == ETL
 
-The ActiveWarehouse plugin does not directly handle Extract-Transform-Load processes, however the ActiveWarehouse ETL gem (installed separately) can help. To install it use:
+The ActiveWarehouse plugin does not directly handle Extract-Transform-Load
+processes, however the ActiveWarehouse ETL gem (installed separately) can help.
+To install it use:
 
   gem install activewarehouse-etl
+
+Once again you should run this command as root or using sudo.
   
-More information on the ETL process can be found at http://activewarehouse.rubyforge.org/etl
+More information on the ETL process can be found at
+http://activewarehouse.rubyforge.org/etl
 
 == Tutorial
 
-A tutorial for ActiveWarehouse is available online at http://anthonyeden.com/2006/12/20/activewarehouse-example-with-rails-svn-logs
+A tutorial for ActiveWarehouse is available online at
+http://anthonyeden.com/2006/12/20/activewarehouse-example-with-rails-svn-logs
+(Note that is is out of date.)
 
-You can also get a demo from the ActiveWarehouse subversion repository. Look in the SVN_ROOT/demo directory.
+You can also get a demo from the ActiveWarehouse subversion repository. Look in
+the SVN_ROOT/demo directory.

data/Rakefile

@@ -68,11 +68,13 @@ spec = Gem::Specification.new do |s|
     ActiveWarehouse extends Rails to provide functionality specific for building data warehouses.
   EOF
 
-  s.add_dependency('rake', '>= 0.7.1')
-  s.add_dependency('activesupport',    '>= 1.3.1.5618')
-  s.add_dependency('activerecord',     '>= 1.14.4.5618')
-  s.add_dependency('actionpack',       '>= 1.12.5.5618')
-  s.add_dependency('rails_sql_views',  '>= 0.1.0')
+  s.add_dependency('rake',                '>= 0.7.1')
+  s.add_dependency('fastercsv',           '>= 1.1.0')
+  s.add_dependency('activesupport',       '>= 1.3.1')
+  s.add_dependency('activerecord',        '>= 1.14.4')
+  s.add_dependency('actionpack',          '>= 1.12.5')
+  s.add_dependency('rails_sql_views',     '>= 0.1.0')
+  s.add_dependency('adapter_extensions',  '>= 0.1.0')
 
   s.rdoc_options << '--exclude' << '.'
   s.has_rdoc = false
@@ -135,4 +137,13 @@ end
 desc "Publish the API documentation"
 task :pdoc => [:rdoc] do 
   Rake::SshDirPublisher.new("aeden@rubyforge.org", "/var/www/gforge-projects/activewarehouse/rdoc", "rdoc").upload
+end
+
+desc "Reinstall the gem from a local package copy"
+task :reinstall => [:package] do
+  windows = RUBY_PLATFORM =~ /mswin/
+  sudo = windows ? '' : 'sudo'
+  gem = windows ? 'gem.bat' : 'gem'
+  `#{sudo} #{gem} uninstall -x -i #{PKG_NAME}`
+  `#{sudo} #{gem} install pkg/#{PKG_NAME}-#{PKG_VERSION}`
 end

data/doc/references.txt

@@ -0,0 +1,4 @@
+The following papers are relevant to building, storing and querying data cubes with large databases.
+
+http://research.microsoft.com/research/pubs/view.aspx?msr_tr_id=MSR-TR-95-22
+http://dbpubs.stanford.edu/pub/showDoc.Fulltext?lang=en&doc=1995-34&format=pdf&compression=&name=1995-34.pdf

data/generators/bridge/templates/migration.rb

@@ -1,8 +1,14 @@
 class <%= migration_name %> < ActiveRecord::Migration
   def self.up
     fields = {
-      # Add bridge attributes here as name => type
-      # Example: :top_flag => :integer
+      # the following are the required bridge table columns for
+      # variable depth hierarchies.  Do not change them unless you know
+      # what you are doing.
+      :parent_id => :integer,
+      :child_id => :integer,
+      :num_levels_from_parent => :integer,
+      :is_bottom => :boolean,
+      :is_top => :boolean
     }
     create_table :<%= table_name %> do |t|
       fields.each do |name,type|
@@ -12,6 +18,7 @@ class <%= migration_name %> < ActiveRecord::Migration
     fields.each do |name,type|
       add_index :<%= table_name %>, name unless type == :text      
     end
+    add_index :<%= table_name %>, [:parent_id, :child_id, :num_levels_from_parent], :unique => true
   end
 
   def self.down

data/generators/bridge/templates/unit_test.rb

@@ -0,0 +1,8 @@
+require File.dirname(__FILE__) + '<%= '/..' * class_nesting_depth %>/../test_helper'
+
+class <%= class_name %>Test < Test::Unit::TestCase
+  # Replace this with your real tests.
+  def test_truth
+    assert true
+  end
+end

data/generators/date_dimension/USAGE

@@ -0,0 +1 @@
+./script/generate date_dimension

data/generators/date_dimension/date_dimension_generator.rb

@@ -0,0 +1,16 @@
+class DateDimensionGenerator < DimensionGenerator
+  attr_accessor :file_name
+  attr_accessor :include_fiscal_year
+  
+  default_options :skip_migration => false
+  
+  def initialize(runtime_args, runtime_options = {})
+    super
+    
+    @name = 'date'
+    @table_name = "#{@name}_dimension"
+    @class_name = "#{@name.camelize}Dimension"
+    @file_name = "#{@class_name.tableize.singularize}"
+    @include_fiscal_year = true # TODO: accept a runtime option to set this
+  end
+end

data/generators/date_dimension/templates/fixture.yml

@@ -0,0 +1,5 @@
+# Read about fixtures at http://ar.rubyonrails.org/classes/Fixtures.html
+first:
+  id: 1
+another:
+  id: 2

data/generators/date_dimension/templates/migration.rb

@@ -0,0 +1,31 @@
+class <%= migration_name %> < ActiveRecord::Migration
+  def self.up
+    create_table :<%= table_name %> do |t|
+      t.column :sql_date, :date, :null => false                             # SQL Date object
+      t.column :calendar_year, :string, :null => false                      # 2005, 2006, 2007, etc.
+      t.column :calendar_quarter, :string, :null => false, :limit => 2      # Q1, Q2, Q3 or Q4
+      t.column :calendar_quarter_number, :integer, :null => false           # 1, 2, 3 or 4
+      t.column :calendar_month_name, :string, :null => false, :limit => 9   # January, February, etc.
+      t.column :calendar_month_number, :integer, :null => false             # 1, 2, 3, ... 12
+      t.column :calendar_week, :string, :null => false, :limit => 2         # 1, 2, 3, ... 52
+      t.column :calendar_week_number, :integer, :null => false              # 1, 2, 3, ... 52
+      t.column :day_number_in_calendar_year, :integer, :null => false       # 1, 2, 3, ... 365
+      t.column :day_number_in_calendar_month, :integer, :null => false      # 1, 2, 3, ... 31
+      t.column :day_in_week, :string, :null => false, :limit => 9           # Monday, Tuesday, etc.
+      <%  if include_fiscal_year -%>
+      t.column :fiscal_year, :string, :null => false
+      t.column :fiscal_quarter, :string, :null => false, :limit => 2
+      t.column :fiscal_quarter_number, :integer, :null => false
+      t.column :fiscal_month_number, :integer, :null => false
+      t.column :fiscal_week, :string, :null => false, :limit => 2
+      t.column :fiscal_week_number, :integer, :null => false
+      t.column :day_number_in_fiscal_year, :integer, :null => false
+      <%  end -%>
+    end
+    # add indexes as required
+  end
+
+  def self.down
+    drop_table :<%= table_name %>
+  end
+end

data/generators/date_dimension/templates/model.rb

@@ -0,0 +1,3 @@
+class <%= class_name %> < ActiveWarehouse::DateDimension
+  
+end

data/generators/date_dimension/templates/unit_test.rb

@@ -0,0 +1,8 @@
+require File.dirname(__FILE__) + '<%= '/..' * class_nesting_depth %>/../test_helper'
+
+class <%= class_name %>Test < Test::Unit::TestCase
+  # Replace this with your real tests.
+  def test_truth
+    assert true
+  end
+end

data/generators/dimension/templates/migration.rb

@@ -1,16 +1,7 @@
 class <%= migration_name %> < ActiveRecord::Migration
   def self.up
-    fields = {
-      # Add dimension attributes here as name => type
-      # Example: :store_name => :string
-    }
     create_table :<%= table_name %> do |t|
-      fields.each do |name,type|
-        t.column name, type
-      end
-    end
-    fields.each do |name,type|
-      add_index :<%= table_name %>, name unless type == :text      
+      
     end
   end
 

data/generators/dimension_view/dimension_view_generator.rb

@@ -21,7 +21,7 @@ class DimensionViewGenerator < Rails::Generator::NamedBase
     # define the query target class and expose its columns as attributes for the view
     @query_target_class = @query_target_class_name.constantize
     @view_attributes = @query_target_class.column_names
-    @view_query = "select * from #{query_target_table_name}"
+    @view_query = "select #{@view_attributes.join(',')} from #{query_target_table_name}"
   end
   
   def manifest
@@ -37,7 +37,7 @@ class DimensionViewGenerator < Rails::Generator::NamedBase
       # Generate the files
       m.template 'model.rb', File.join('app/models', class_path, "#{file_name}.rb")
       m.template 'unit_test.rb', File.join('test/unit', class_path, "#{file_name}_test.rb")
-      m.template 'fixture.yml', File.join('test/fixtures', class_path, "#{table_name}.yml")
+      #m.template 'fixture.yml', File.join('test/fixtures', class_path, "#{table_name}.yml")
       
       # Generate the migration unless :skip_migration option is specified
       unless options[:skip_migration]

data/generators/dimension_view/templates/migration.rb

@@ -1,7 +1,13 @@
 class <%= migration_name %> < ActiveRecord::Migration
   def self.up
-    create_view "<%= view_name %>", <%= view_query do |t|
-      <%= view_attributes %>
+    create_view "<%= view_name %>", "<%= view_query %>" do |t|
+      <%- view_attributes.each do |view_attribute| -%>
+      <%-  if view_attribute == 'id' -%>
+      t.column :id
+      <%-  else -%>
+      t.column :<%= name %>_<%= view_attribute %>
+      <%-  end -%>
+      <%-  end -%>
     end
   end
 

data/generators/fact/templates/migration.rb

@@ -3,6 +3,8 @@ class <%= migration_name %> < ActiveRecord::Migration
     create_table :<%= table_name %> do |t|
       
     end
+    # you should add indexes for each foreign key, but don't add
+    # the foreign key itself unless you really know what you are doing.
   end
 
   def self.down

data/generators/time_dimension/USAGE

@@ -0,0 +1 @@
+./script/generate time_dimension

data/generators/time_dimension/templates/fixture.yml

@@ -0,0 +1,5 @@
+# Read about fixtures at http://ar.rubyonrails.org/classes/Fixtures.html
+first:
+  id: 1
+another:
+  id: 2

data/generators/time_dimension/templates/migration.rb

@@ -0,0 +1,12 @@
+class <%= migration_name %> < ActiveRecord::Migration
+  def self.up
+    create_table :<%= table_name %> do |t|
+      t.column :hour_of_day, :integer, :null => false
+      t.column :minute_of_hour, :integer, :null => false
+    end
+  end
+
+  def self.down
+    drop_table :<%= table_name %>
+  end
+end

data/generators/time_dimension/templates/model.rb

@@ -0,0 +1,3 @@
+class <%= class_name %> < ActiveWarehouse::Dimension
+  
+end

data/generators/time_dimension/templates/unit_test.rb

@@ -0,0 +1,8 @@
+require File.dirname(__FILE__) + '<%= '/..' * class_nesting_depth %>/../test_helper'
+
+class <%= class_name %>Test < Test::Unit::TestCase
+  # Replace this with your real tests.
+  def test_truth
+    assert true
+  end
+end

data/generators/time_dimension/time_dimension_generator.rb

@@ -0,0 +1,14 @@
+class TimeDimensionGenerator < DimensionGenerator
+  attr_accessor :file_name
+  
+  default_options :skip_migration => false
+  
+  def initialize(runtime_args, runtime_options = {})
+    super
+    
+    @name = 'date'
+    @table_name = "#{@name}_dimension"
+    @class_name = "#{@name.camelize}Dimension"
+    @file_name = "#{@class_name.tableize.singularize}"
+  end
+end

data/lib/active_warehouse.rb

@@ -35,7 +35,7 @@ end
 
 unless defined?(ActiveSupport)
   begin
-    $:.unshift(File.dirname(__FILE__) + "/../../activesupport/lib")  
+    $:.unshift(File.dirname(__FILE__) + "/../../activesupport/lib")
     require 'active_support'  
   rescue LoadError
     gem 'activesupport'
@@ -62,18 +62,29 @@ unless defined?(ActionView)
   end
 end
 
+require 'fastercsv'
+require 'fileutils'
+require 'adapter_extensions'
+
 # Require 1.1.6 compatibility code if necessary
 require 'active_record/version'
 if ActiveRecord::VERSION::MAJOR < 1 || ActiveRecord::VERSION::MINOR < 15
   require 'active_warehouse/compat/compat'
 end
 
+require 'active_warehouse/ordered_hash'
+require 'active_warehouse/field'
+require 'active_warehouse/aggregate_field'
+require 'active_warehouse/calculated_field'
 require 'active_warehouse/version'
 require 'active_warehouse/core_ext'
-require 'active_warehouse/aggregate'
+require 'active_warehouse/prejoin_fact'
 require 'active_warehouse/fact'
+require 'active_warehouse/bridge'
 require 'active_warehouse/dimension'
 require 'active_warehouse/cube'
+require 'active_warehouse/cube_query_result'
+require 'active_warehouse/aggregate'
 require 'active_warehouse/report'
 require 'active_warehouse/view'
 require 'active_warehouse/builder'

data/lib/active_warehouse/aggregate.rb

@@ -1,260 +1,61 @@
-module ActiveWarehouse #:nodoc:
-  # An aggreate within a cube used to store calculated values. Each aggregate will contain values for a dimension pair, 
-  # down each of the dimension hierarchies
-  class Aggregate < ActiveRecord::Base
-    class << self
-      attr_accessor :name, :cube, :dimension1, :dimension2, :dimension1_hierarchy_name, :dimension2_hierarchy_name
-      
-      # Get the table name for the aggregate
-      def table_name
-        name = self.name.demodulize.underscore
-        set_table_name(name)
-        name
-      end
-      
-      # Returns the aggregate ID
-      def aggregate_id
-        table_name =~ /(\d+)$/
-        $1.to_i
-      end
-      
-      # Returns the AggregateMetaData instance associated with this aggregate
-      def meta_data
-        AggregateMetaData.find(aggregate_id)
-      end
-      
-      # Return true if the aggregate needs to be rebuilt
-      def needs_rebuild?(last_build=nil)
-        return true if meta_data.populated_at.nil?
-        return true if last_build && (meta_data.populated_at < last_build)
-        return false
-      end
-      
-      # Return a key for the aggregate
-      def key(dimension1, dimension1_hierarchy, dimension2, dimension2_hierarchy)
-        AggregateKey.new(dimension1, dimension1_hierarchy, dimension2, dimension2_hierarchy)
-      end
-      
-      # Create the aggregate table if required. Set force option to true to force creation of the table
-      # if it already exists
-      def create_storage_table(force=false)
-        connection.drop_table(table_name) if force and table_exists?
-        if !table_exists?
-          connection.create_table(table_name, :id => false) do |t|
-            t.column :dimension1_path, :string
-            t.column :dimension1_stage, :integer
-            t.column :dimension2_path, :string
-            t.column :dimension2_stage, :integer
-            cube.fact_class.aggregate_fields.each do |field|
-              #options = cube.fact_class.aggregate_field_options[field]
-              col = cube.fact_class.columns_hash[field.to_s]
-              t.column field, col.type if col
-            end
-          end
-          connection.add_index(table_name, :dimension1_path)
-          connection.add_index(table_name, :dimension1_stage)
-          connection.add_index(table_name, :dimension2_path)
-          connection.add_index(table_name, :dimension2_stage)
-        end
-      end
-      
-      # Populate the aggregate table
-      def populate
-        # create the storage table if necessary
-        create_storage_table
-        
-        #puts "Populating aggregate table #{table_name}"
-        # clear out the current data
-        #connection.execute("TRUNCATE TABLE #{table_name}") #TODO: make this generic to support all databases
-        delete_all
-        
-        $first = false
-        
-        # aggregate the data for the two dimensions
-        fact_class = cube.fact_class
-        dim1 = Dimension.class_name(dimension1).constantize
-        dim2 = Dimension.class_name(dimension2).constantize
-        dim1_stage_path = []
-        dim1.hierarchy(meta_data.dimension1_hierarchy.to_sym).each_with_index do |dim1_stage_name, dim1_stage_level|
-          dim1_stage_path << dim1_stage_name
-          dim2_stage_path = []
-          dim2.hierarchy(meta_data.dimension2_hierarchy.to_sym).each_with_index do |dim2_stage_name, dim2_stage_level|
-            dim2_stage_path << dim2_stage_name
-            
-            stmt, fields = build_query(fact_class, dim1, dim1_stage_path, dim2, dim2_stage_path)
-            
-            puts "\nSTMT: #{stmt}" if $first
-            
-            # Get the facts and aggregate them
-            # TODO: replace with select_all
-            fact_class.connection.select_all(stmt).each do |row|
-              require 'pp'
-              pp row if $first
-              dim1_value = []
-              dim1_stage_path.each do |v|
-                dim1_value << row["#{v}"]
-              end
-              dim2_value = []
-              dim2_stage_path.each do |v|
-                dim2_value << row["#{v}"]
-              end
-              
-              agg_instance = new
-              agg_instance.dimension1_path = dim1_value.join(':')
-              agg_instance.dimension1_stage = dim1_stage_level
-              agg_instance.dimension2_path = dim2_value.join(':')
-              agg_instance.dimension2_stage = dim2_stage_level
-              
-              puts "DIM1_PATH: #{agg_instance.dimension1_path}" if $first
-              puts "DIM2_PATH: #{agg_instance.dimension2_path}" if $first
-              
-              
-              pp fields if $first
-              fields.each do |field|
-                # do the average here
-                puts "setting field #{field}, value is #{row[field.to_s]}" if $first
-                agg_instance.send("#{field}=".to_sym, row[field.to_s])
-              end
-              agg_instance.save!
-              
-              meta_data.update_attribute(:populated_at, Time.now)
-              
-              $first = false
-            end
-          end
-        end
-      end
-      
-      # Build the aggregation query for the given dimensions and stage paths
-      def build_query(fact_class, dim1, dim1_stage_path, dim2, dim2_stage_path)
-        dim1_group = dim1_stage_path.collect { |p| "d1.#{p}"}.join(", ")
-        dim2_group = dim2_stage_path.collect { |p| "d2.#{p}"}.join(", ")
-
-        # Set up the find options
-        fact_find_options = {}
-        fact_find_options[:group] = "#{dim1_group}, #{dim2_group}"
-        fact_find_options[:joins] = "join #{dim1.table_name} d1 on f.#{dim1.foreign_key} = d1.id"
-        fact_find_options[:joins] << " join #{dim2.table_name} d2 on f.#{dim2.foreign_key} = d2.id"
-
-        # Build the 'select' part of the query
-        # denominator = nil
-        fields = []
-        fact_select = ["#{dim1_group}, #{dim2_group}"]
-        fact_class.aggregate_fields.each do |field_name|
-          options = fact_class.aggregate_field_options[field_name]
-          fields << field_name
-          
-          options[:type] ||= :sum
-          case options[:type]
-          when :sum
-            fact_select << " sum(f.#{field_name}) as #{field_name}"
-          when Hash
-            if options[:type][dim1.sym] == :average && options[:type][dim2.sym] == :average
-              # I believe this is a special case, but I'm not sure how yet. If both dimensions are defined
-              # averages then perhaps that value cannot be calculated at all. TODO: research
-            else
-              fact_select << " sum(f.#{field_name}) as #{field_name}"
-            end
-          else
-            raise "Unsupported aggregate type: #{options[:type]}"
-          end
-        end
-        fact_find_options[:select] = fact_select.join(',')
+# Source file which defines the ActiveWarehouse::Aggregate module and imports
+# the aggregate implementations.
 
-        # put the SQL statement together
-        stmt = "select #{fact_find_options[:select]} from "
-        stmt << "#{fact_class.table_name} f #{fact_find_options[:joins]} "
-        stmt << "group by #{fact_find_options[:group]} "
-
-        return stmt.strip, fields
-      end
-      
-    end
-    
-    public
-    # Clone and reset at the same time
-    def clone_and_reset
-      o = clone
-      o.reset
-      o
-    end
-    
-    def non_data_fields
-      ['dimension1_path','dimension1_stage','dimension2_path','dimension2_stage']
-    end
-    
-    def data_fields
-      fields = []
-      self.class.columns.each do |column|
-        unless non_data_fields.include?(column.name)
-          fields << column.name
-        end
-      end
-      fields
-    end
-    
-    # Reset the aggregate
-    def reset
-      self.class.columns.each do |column|
-        unless non_data_fields.include?(column.name)
-          value = column.number? ? 0 : 'None'
-          send("#{column.name}=".to_sym, value)
-        end
-      end
-    end
-  end
-  
-  # ActiveRecord object which stores meta data about the aggregate
-  class AggregateMetaData < ActiveRecord::Base
-    # Build the underlying table. Set force to true to force the build of the table
-    def self.build_table(force=false)
-      connection.drop_table(table_name) if force and table_exists?
-      if !table_exists?
-        connection.create_table(table_name) do |t|
-          t.column :cube_name, :string
-          t.column :dimension1, :string
-          t.column :dimension1_hierarchy, :string
-          t.column :dimension2, :string
-          t.column :dimension2_hierarchy, :string
-          t.column :created_at, :datetime
-          t.column :populated_at, :datetime
+module ActiveWarehouse #:nodoc:
+  # This module contains classes which handle aggregation of cube data using
+  # various algorithms
+  module Aggregate
+    # Base class for aggregate implementations
+    class Aggregate
+      
+      # Reader for the cube class
+      attr_reader :cube_class
+      
+      # Initialize the aggregate for the given cube class
+      def initialize(cube_class)
+        @cube_class = cube_class
+      end
+      
+      protected
+      # Get the connection to use for SQL execution
+      def connection
+        cube_class.connection
+      end
+      
+      # Convenience accessor to get the cube's fact class. Delegates to the
+      # cube class.
+      def fact_class
+        cube_class.fact_class
+      end
+      
+      # Parse the query args and return an options hash.
+      def parse_query_args(*args)
+        options = {}
+        if args.length == 1
+          options = args[0]
+        elsif args.length >= 4
+          options[:column_dimension_name] = args[0]
+          options[:column_hierarchy_name] = args[1]
+          options[:row_dimension_name] = args[2]
+          options[:row_hierarchy_name] = args[3]
+          options[:conditions] = args[4] if args.length >= 5
+          options[:cstage] = args[5] if args.length >= 6
+          options[:rstage] = args[6] if args.length >= 7
+          options[:filters] = args[7] if args.length >= 8
+          options.merge!(args[8]) if args.length >= 9
+        else
+          raise ArgumentError, "The query method accepts either 1 Hash (new style) or 4 to 8 arguments (old style)"
         end
-        connection.add_index table_name, :cube_name
+        options
       end
     end
-    def key
-      Aggregate.key(dimension1, dimension1_hierarchy, dimension2, dimension2_hierarchy)
-    end
-  end
-  
-  # Key for aggregate caching
-  class AggregateKey
-    attr_reader :dimension1, :dimension1_hierarchy, :dimension2, :dimension2_hierarchy
-    
-    def initialize(dimension1, dimension1_hierarchy, dimension2, dimension2_hierarchy)
-      @dimension1 = dimension1
-      @dimension1_hierarchy = dimension1_hierarchy
-      @dimension2 = dimension2
-      @dimension2_hierarchy = dimension2_hierarchy
-    end
-    
-    def ==(o)
-      o.instance_of?(self.class) and (o.to_s == to_s or o.to_s = to_rs)
-    end
-    
-    def hash
-      to_s.hash
-    end
-    
-    def to_s
-      "#{@dimension1}.#{@dimension1_hierarchy}.#{@dimension2}.#{@dimension2_hierarchy}"
-    end
-    
-    # Return the "reveresed" version of this key String representation
-    def to_rs
-      "#{@dimension2}.#{@dimension2_hierarchy}.#{@dimension1}.#{@dimension1_hierarchy}"
-    end
   end
 end
 
-ActiveWarehouse::AggregateMetaData.build_table
+require 'active_warehouse/aggregate/no_aggregate'
+require 'active_warehouse/aggregate/rolap_common'
+require 'active_warehouse/aggregate/pipelined_rolap_aggregate'
+require 'active_warehouse/aggregate/rolap_aggregate'
+require 'active_warehouse/aggregate/dwarf_common'
+require 'active_warehouse/aggregate/dwarf_aggregate'
+require 'active_warehouse/aggregate/pid_aggregate'