chicagowarehouse 0.5.1 → 0.6.0

checksums.yaml

@@ -1,15 +1,15 @@
 ---
 !binary "U0hBMQ==":
   metadata.gz: !binary |-
-    OTZhYTliODU1NjljZmRkNGE4YWZkOGFjMDcxMGMwMjRhMWNlY2E2Mw==
+    NmI5NTE0ZmI4MTJhZDc3MjE2YTkzMTM5NWZhY2M5MDIxMTBjODI1Yg==
   data.tar.gz: !binary |-
-    ZDU5Yjc3YmRhZjc2MjEyMTg1MDExODE3N2U0MzE4MGM2NDQ1YjkyNw==
+    MGM2YjI2ZWFkNzhlMzI3NzM0MmY0YjgyMGQ5MjdhYzYxNGQ3ZWM1MQ==
 SHA512:
   metadata.gz: !binary |-
-    YTViYWJiNDk2NDMwMzkyODNmZWRmOTBhZmEyMDE1ODBmODU3MDY4MTg4MmRl
-    MGQ1NjMwMDU0OTRmNjkzZmQxN2E5M2QwY2U2ZjU2NjQ1NTE3ZTY2ZTVmMGQw
-    MTczN2U0N2VhMmU3MGJkZGE1NjJkMmNiMGEzYTVlN2ViYzFjN2I=
+    Y2MwMDQ2M2M4NTdiMzE0MzY0MjAwOWU4YmYwYzI5ZjI0NDdhYTU2ZTU3NGQw
+    OThiZjZiZjFmYWYyOGZmYzdkY2EyZjZiYzRjYjE2OWI5Y2ZlNmZiZDNjOWQy
+    NDIwZTZhNGMwYjAxNWJkMTJjZmQ3NDAzNTdkMWVjZDlmMjgxMWQ=
   data.tar.gz: !binary |-
-    MzZhNzJiY2FjY2Y0ZGE1N2IzZTMyOGIxNGQ0MTQ3YTZkY2ExMjhjNTY2MDZl
-    MTU3YWZlOTA2YTExNzI0NDFhZTRkYTBhNTJiMTZlNTcwMjVlYWZiZjMwOTY5
-    ZjdkMDZjYTI5ZjFjNTViY2Y0ZGU3N2NiNDQzMTYzMzgwM2Q0YTE=
+    NzQ1YTA5NjM4NWVlMmU4YmJhMTM3YjhlMzQ5MzgwMWY4YmEwZmU2MTliYzFk
+    OWRmNDAwMDQwNWY1NWFkMzVkODlhYjZiM2U1MGNjZGI0OWQ3ZTViMWQ4ZDg1
+    NDYyMzBjZDA5NzU4NDBkOGE1NTQ2OGM1NGJlZDQ0ZTNkMzJkZGM=

data/Gemfile

@@ -1,7 +1,7 @@
 source "http://rubygems.org"
 
 gem "sequel", ">= 4.7"
-gem "sequel_migration_builder", ">= 0.3.2"
+gem "sequel_migration_builder", ">= 0.4.0"
 gem "chronic"
 
 group :development do
@@ -10,7 +10,6 @@ group :development do
   gem "rspec", "~> 2.0"
   gem "bundler"
   gem "jeweler"
-  gem "rcov", :platform => :ruby_18
   gem "simplecov", :platform => :ruby_19
   gem "flog"
   gem "timecop"

data/Rakefile

@@ -14,7 +14,7 @@ require 'rake'
 require 'jeweler'
 Jeweler::Tasks.new do |gem|
   gem.name              = "chicagowarehouse"
-  gem.version           = "0.5.1"
+  gem.version           = "0.6.0"
   gem.summary           = "Ruby Data Warehousing"
   gem.description       = "Simple Data Warehouse toolkit for ruby"
   gem.author            = "Roland Swingler"

data/chicagowarehouse.gemspec

@@ -2,16 +2,16 @@
 # DO NOT EDIT THIS FILE DIRECTLY
 # Instead, edit Jeweler::Tasks in Rakefile, and run 'rake gemspec'
 # -*- encoding: utf-8 -*-
-# stub: chicagowarehouse 0.5.1 ruby lib
+# stub: chicagowarehouse 0.6.0 ruby lib
 
 Gem::Specification.new do |s|
   s.name = "chicagowarehouse"
-  s.version = "0.5.1"
+  s.version = "0.6.0"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.require_paths = ["lib"]
   s.authors = ["Roland Swingler"]
-  s.date = "2014-07-03"
+  s.date = "2014-09-01"
   s.description = "Simple Data Warehouse toolkit for ruby"
   s.email = "roland.swingler@gmail.com"
   s.extra_rdoc_files = [
@@ -30,12 +30,13 @@ Gem::Specification.new do |s|
     "lib/chicago/core_ext/hash.rb",
     "lib/chicago/core_ext/sequel/dataset.rb",
     "lib/chicago/data/month.rb",
+    "lib/chicago/database/concrete_schema_strategies.rb",
     "lib/chicago/database/constants.rb",
     "lib/chicago/database/dataset_builder.rb",
     "lib/chicago/database/filter.rb",
+    "lib/chicago/database/index_generator.rb",
     "lib/chicago/database/migration_file_writer.rb",
     "lib/chicago/database/schema_generator.rb",
-    "lib/chicago/database/type_converters.rb",
     "lib/chicago/database/value_parser.rb",
     "lib/chicago/errors.rb",
     "lib/chicago/query.rb",
@@ -57,7 +58,7 @@ Gem::Specification.new do |s|
     "lib/chicago/schema/table.rb",
     "lib/chicago/star_schema.rb",
     "spec/data/month_spec.rb",
-    "spec/database/db_type_converter_spec.rb",
+    "spec/database/concrete_schema_strategies.rb",
     "spec/database/migration_file_writer_spec.rb",
     "spec/database/schema_generator_spec.rb",
     "spec/db_connections.yml.dist",
@@ -90,41 +91,38 @@ Gem::Specification.new do |s|
 
     if Gem::Version.new(Gem::VERSION) >= Gem::Version.new('1.2.0') then
       s.add_runtime_dependency(%q<sequel>, [">= 4.7"])
-      s.add_runtime_dependency(%q<sequel_migration_builder>, [">= 0.3.2"])
+      s.add_runtime_dependency(%q<sequel_migration_builder>, [">= 0.4.0"])
       s.add_runtime_dependency(%q<chronic>, [">= 0"])
       s.add_development_dependency(%q<mysql2>, [">= 0"])
       s.add_development_dependency(%q<yard>, [">= 0"])
       s.add_development_dependency(%q<rspec>, ["~> 2.0"])
       s.add_development_dependency(%q<bundler>, [">= 0"])
       s.add_development_dependency(%q<jeweler>, [">= 0"])
-      s.add_development_dependency(%q<rcov>, [">= 0"])
       s.add_development_dependency(%q<simplecov>, [">= 0"])
       s.add_development_dependency(%q<flog>, [">= 0"])
       s.add_development_dependency(%q<timecop>, [">= 0"])
     else
       s.add_dependency(%q<sequel>, [">= 4.7"])
-      s.add_dependency(%q<sequel_migration_builder>, [">= 0.3.2"])
+      s.add_dependency(%q<sequel_migration_builder>, [">= 0.4.0"])
       s.add_dependency(%q<chronic>, [">= 0"])
       s.add_dependency(%q<mysql2>, [">= 0"])
       s.add_dependency(%q<yard>, [">= 0"])
       s.add_dependency(%q<rspec>, ["~> 2.0"])
       s.add_dependency(%q<bundler>, [">= 0"])
       s.add_dependency(%q<jeweler>, [">= 0"])
-      s.add_dependency(%q<rcov>, [">= 0"])
       s.add_dependency(%q<simplecov>, [">= 0"])
       s.add_dependency(%q<flog>, [">= 0"])
       s.add_dependency(%q<timecop>, [">= 0"])
     end
   else
     s.add_dependency(%q<sequel>, [">= 4.7"])
-    s.add_dependency(%q<sequel_migration_builder>, [">= 0.3.2"])
+    s.add_dependency(%q<sequel_migration_builder>, [">= 0.4.0"])
     s.add_dependency(%q<chronic>, [">= 0"])
     s.add_dependency(%q<mysql2>, [">= 0"])
     s.add_dependency(%q<yard>, [">= 0"])
     s.add_dependency(%q<rspec>, ["~> 2.0"])
     s.add_dependency(%q<bundler>, [">= 0"])
     s.add_dependency(%q<jeweler>, [">= 0"])
-    s.add_dependency(%q<rcov>, [">= 0"])
     s.add_dependency(%q<simplecov>, [">= 0"])
     s.add_dependency(%q<flog>, [">= 0"])
     s.add_dependency(%q<timecop>, [">= 0"])

data/lib/chicago.rb

@@ -12,7 +12,8 @@ require 'chicago/data/month'
 
 require 'chicago/star_schema'
 require 'chicago/database/constants'
-require 'chicago/database/type_converters'
+require 'chicago/database/index_generator'
+require 'chicago/database/concrete_schema_strategies'
 require 'chicago/database/migration_file_writer'
 require 'chicago/database/schema_generator'
 require 'chicago/query'

data/lib/chicago/database/concrete_schema_strategies.rb

@@ -0,0 +1,153 @@
+module Chicago
+  module Database
+    # Generic database strategy.
+    #
+    # This supplements Sequel's type conversion strategy rather than
+    # replaces it, so +:boolean+ will still return +:boolean+ rather
+    # than +tinyint(1)+ in the case of mysql.
+    class ConcreteSchemaStrategy
+      # Factory method that returns an appropriate type conversion
+      # stratgey for the given database.
+      #
+      # If a database-specific strategy cannot be found, returns a
+      # generic strategy.
+      #
+      # @return [ConcreteSchemaStrategy]
+      def self.for_db(db)
+        if db.database_type == :mysql
+          MysqlStrategy.new
+        elsif db.database_type == :postgres && db.opts[:adapter] == "redshift"
+          RedshiftStrategy.new
+        else
+          self.new
+        end
+      end
+
+      def migration_options
+        {}
+      end
+
+      def column_hash(column)
+        hsh = column.to_hash.merge(:column_type => db_type(column))
+        hsh.delete(:elements) if hsh.has_key?(:elements)
+        hsh
+      end
+
+      # Returns the indexes for the given table.
+      def indexes(table)
+        IndexGenerator.new(table).indexes
+      end
+
+      # Returns a db type given a column definition
+      #
+      # @return [Symbol]
+      def db_type(column)
+        case column.column_type
+        when :integer then integer_type(column.min, column.max)
+        when :string  then string_type(column.min, column.max)
+        when :money   then :decimal
+        when :percent then :decimal
+        else
+          column.column_type
+        end
+      end
+
+      # Returns sequel table options for a dimension or fact table.
+      #
+      # None by default, but database-specific subclasses may
+      # override this.
+      #
+      # @return [Hash]
+      def table_options
+        {}
+      end
+
+      # Returns a database type for a string column.
+      #
+      # @return [Symbol]
+      def string_type(min, max)
+        min && max && min == max ? :char : :varchar 
+      end
+      
+      # Returns a database integer column type, big enough to fit
+      # values between min and max, or integer if a specific type
+      # cannot be found.
+      #
+      # @return [Symbol]
+      # @raise an ArgumentError if min or max is too large for a
+      #   single database column.
+      def integer_type(min, max)
+        if min && max && in_numeric_range?(min, max, SMALL_INT_MAX)
+          :smallint
+        else
+          :integer
+        end
+      end
+
+      protected
+
+      def in_numeric_range?(min, max, unsigned_limit)
+        signed_limit = (unsigned_limit + 1) / 2
+        (min >= -signed_limit && max <= signed_limit - 1)  ||  (min >= 0 && max <= unsigned_limit)
+      end
+    end
+
+    # Redshift-specific database schema strategy
+    class RedshiftStrategy < ConcreteSchemaStrategy
+      def migration_options
+        {:separate_alter_table_statements => true, :immutable_columns => true}
+      end
+
+      def column_hash(column)
+        hsh = super(column)
+
+        if column.column_type == :string && hsh[:size]
+          # Redshift column sizes are in bytes, not characters, so
+          # increase to 4 bytes per-char for UTF-8 reasons.
+          hsh[:size] *= 4
+        end
+
+        hsh
+      end
+
+      # Redshift does not support indexes, so do not output any.
+      def indexes(table)
+        []
+      end
+    end
+
+    # MySql-specific database schema strategy
+    class MysqlStrategy < ConcreteSchemaStrategy
+      def column_hash(column)
+        column.to_hash.merge :column_type => db_type(column)
+      end
+
+      def db_type(column)
+        return :enum if column.elements && column.elements.size < 65_536
+        super(column)
+      end
+
+      # Returns table options for a dimension or fact table.
+      #
+      # Dimension tables are defined as MyISAM tables in MySQL.
+      def table_options
+        {:engine => "myisam"}
+      end
+
+      def integer_type(min, max)
+        return :integer unless min && max
+
+        case
+        when in_numeric_range?(min, max, TINY_INT_MAX)   then :tinyint
+        when in_numeric_range?(min, max, SMALL_INT_MAX)  then :smallint
+        when in_numeric_range?(min, max, MEDIUM_INT_MAX) then :mediumint
+        when in_numeric_range?(min, max, INT_MAX)        then :integer
+        when in_numeric_range?(min, max, BIG_INT_MAX)    then :bigint
+        else
+          raise ArgumentError.new("#{min} is too small or #{max} is too large for a single column")
+        end
+      end
+    end
+  end
+end
+

data/lib/chicago/database/index_generator.rb

@@ -0,0 +1,35 @@
+module Chicago
+  module Database
+    class IndexGenerator
+      def initialize(table)
+        @table = table
+      end
+
+      def indexes
+        indexes = @table.columns.select(&:indexed?).inject({}) do |hsh, d|
+          hsh.merge("#{d.name}_idx".to_sym => {
+                      :columns => d.database_name,
+                      :unique => d.unique?})
+        end
+        indexes.merge!(natural_key_index) if @table.natural_key
+        indexes.merge!(:_inserted_at_idx => {:columns => :_inserted_at, :unique => false})
+        indexes
+      end
+
+      def natural_key_index
+        {
+          "#{@table.natural_key.first}_idx".to_sym => {
+            :columns => natural_key_index_columns,
+            :unique => true
+          }
+        }
+      end
+
+      def natural_key_index_columns
+        @table.natural_key.map do |name|
+          @table[name].database_name rescue raise MissingDefinitionError.new("Column #{name} is not defined in #{@table.name}")
+        end
+      end
+    end
+  end
+end

data/lib/chicago/database/migration_file_writer.rb

@@ -4,30 +4,27 @@ module Chicago
   module Database
     # Writes Sequel migrations for the star schema
     class MigrationFileWriter
-      # Creates a new migration file writer, given a Sequel::Database
-      # connection and a directory. If the directory does not exist, an
-      # error will be raised.
-      def initialize(db, migration_directory)
-        @db = db
-        @migration_directory = migration_directory
-      end
-
       # Writes the migration file necessary for all defined facts and
       # dimensions.
-      def write_migration_file(schema)
-        @file = nil
-        type_converter = TypeConverters::DbTypeConverter.for_db(@db)
-        tables = SchemaGenerator.new(type_converter).traverse(schema)
+      def write_migration_file(db, schema, directory, generate_key_tables=true)
+        schema_strategy = ConcreteSchemaStrategy.for_db(db)
+        tables = SchemaGenerator.new(schema_strategy, generate_key_tables).traverse(schema)
 
-        File.open(migration_file, "w") do |fh|
-          fh.write Sequel::MigrationBuilder.new(@db).generate_migration(tables)
+        File.open(migration_file(directory), "w") do |fh|
+          fh.write Sequel::MigrationBuilder.new(db, schema_strategy.migration_options).
+            generate_migration(tables)
         end
       end
 
       # Returns the path the migration file has been written to.
-      def migration_file
-        @file ||= File.join(@migration_directory, 
-                            "#{Time.now.strftime("%Y%m%d%H%M%S")}_auto_migration.rb")
+      def migration_file(directory)
+        File.join(directory, migration_file_name)
+      end
+
+      private
+
+      def migration_file_name
+        @migration_file_name ||= "#{Time.now.strftime("%Y%m%d%H%M%S")}_auto_migration.rb"
       end
     end
   end

data/lib/chicago/database/schema_generator.rb

@@ -5,10 +5,15 @@ module Chicago
     class SchemaGenerator
       attr_writer :type_converter
       
-      def initialize(type_converter)
+      def initialize(type_converter, generate_key_tables=true)
         @type_converter = type_converter
+        @generate_key_tables = generate_key_tables
       end
       
+      def generate_key_tables?
+        @generate_key_tables
+      end
+
       def traverse(schema)
         schema.tables.inject({}) {|hsh,t| hsh.merge(t.visit(self)) }
       end
@@ -18,12 +23,13 @@ module Chicago
       end
 
       def visit_dimension(dimension)
-        {dimension.table_name => basic_table(dimension)}.
-          merge!(key_table(dimension))
+        hash = {dimension.table_name => basic_table(dimension)}
+        hash.merge!(key_table(dimension)) if generate_key_tables?
+        hash
       end
 
       def visit_column(column)
-        column.to_hash.merge :column_type => @type_converter.db_type(column)
+        @type_converter.column_hash(column)
       end
 
       alias :visit_measure :visit_column
@@ -35,7 +41,7 @@ module Chicago
         t = {
           :primary_key => [:id],
           :table_options => @type_converter.table_options,
-          :indexes => indexes(table),
+          :indexes => @type_converter.indexes(table),
           :columns => [{
                          :name => :id,
                          :column_type => :integer,
@@ -75,42 +81,6 @@ module Chicago
           }
         }
       end
-      
-      def indexes(table)
-        IndexGenerator.new(table).indexes
-      end
-    end
-
-    class IndexGenerator
-      def initialize(table)
-        @table = table
-      end
-
-      def indexes
-        indexes = @table.columns.select(&:indexed?).inject({}) do |hsh, d|
-          hsh.merge("#{d.name}_idx".to_sym => {
-                      :columns => d.database_name,
-                      :unique => d.unique?})
-        end
-        indexes.merge!(natural_key_index) if @table.natural_key
-        indexes.merge!(:_inserted_at_idx => {:columns => :_inserted_at, :unique => false})
-        indexes
-      end
-
-      def natural_key_index
-        {
-          "#{@table.natural_key.first}_idx".to_sym => {
-            :columns => natural_key_index_columns,
-            :unique => true
-          }
-        }
-      end
-
-      def natural_key_index_columns
-        @table.natural_key.map do |name|
-          @table[name].database_name rescue raise MissingDefinitionError.new("Column #{name} is not defined in #{@table.name}")
-        end
-      end
     end
   end
 end

data/lib/chicago/rake_tasks.rb

@@ -20,10 +20,12 @@ module Chicago
   #
   # @api public
   class RakeTasks < Rake::TaskLib
-    def initialize(db, schema)
-      @migration_dir = "migrations"
-      @db = db
+    def initialize(schema, options)
       @schema = schema
+      @base_migration_dir = options[:migration_directory] ||= "migrations"
+      @staging_db = options[:staging_db] or raise ArgumentError.new("staging_db option must be provided.")
+      @presentation_db = options[:presentation_db]
+
       define
     end
 
@@ -36,15 +38,33 @@ module Chicago
         task :create_null_records do
           # TODO: replace this with proper logging.
           warn "Loading NULL records."
-          @schema.dimensions.each {|dimension| dimension.create_null_records(@db) }
+          @schema.dimensions.each do |dimension| 
+            dimension.create_null_records(@db) 
+          end
         end
 
         desc "Writes a migration file to change the database based on defined Facts & Dimensions"
         task :write_migrations do
-          Database::MigrationFileWriter.new(@db, @migration_dir).
-            write_migration_file(@schema)
+          writer = Database::MigrationFileWriter.new
+          writer.write_migration_file(@staging_db, @schema,
+                                      staging_directory)
+
+          if @presentation_db
+            writer.write_migration_file(@presentation_db, @schema, 
+                                        presentation_directory, false)
+          end
         end
       end
     end
+
+    private
+
+    def staging_directory
+      File.join(@base_migration_dir, "staging")
+    end
+
+    def presentation_directory
+      File.join(@base_migration_dir, "presentation")
+    end
   end
 end

data/spec/database/{db_type_converter_spec.rb → concrete_schema_strategies.rb}

@@ -46,15 +46,22 @@ describe "DbTypeConverter.for_db" do
   it "should return a type converter specific to MySQL if the database type is :mysql" do
     @mock_db.should_receive(:database_type).and_return(:mysql)
 
-    converter = Database::TypeConverters::DbTypeConverter.for_db(@mock_db)
-    converter.should be_kind_of(Database::TypeConverters::MysqlTypeConverter)
+    converter = Database::ConcreteSchemaStrategy.for_db(@mock_db)
+    converter.should be_kind_of(Database::MysqlStrategy)
+  end
+
+  it "should return a type converter specific to Redshift if the database type is :mysql" do
+    @mock_db.stub(:database_type => :postgres, :opts => {:adapter => 'redshift'})
+
+    converter = Database::ConcreteSchemaStrategy.for_db(@mock_db)
+    converter.should be_kind_of(Database::RedshiftStrategy)
   end
 
   it "should return a generic type converter for an unknown database type" do
-    @mock_db.should_receive(:database_type).and_return(:foodb)
+    @mock_db.stub(:database_type => :foodb)
 
-    converter = Database::TypeConverters::DbTypeConverter.for_db(@mock_db)
-    converter.should be_kind_of(Database::TypeConverters::DbTypeConverter)
+    converter = Database::ConcreteSchemaStrategy.for_db(@mock_db)
+    converter.should be_kind_of(Database::ConcreteSchemaStrategy)
   end
 end
 
@@ -62,7 +69,7 @@ describe "Generic DbTypeConverter" do
   it_behaves_like "All DB type converters"
 
   before :each do
-    @tc = Database::TypeConverters::DbTypeConverter.new
+    @tc = Database::ConcreteSchemaStrategy.new
   end
 
   { :smallint  => [-32768, 32767],
@@ -76,11 +83,11 @@ describe "Generic DbTypeConverter" do
   end
 end
 
-describe Chicago::Database::TypeConverters::MysqlTypeConverter do
+describe Chicago::Database::MysqlStrategy do
   it_behaves_like "All DB type converters"
 
   before :each do
-    @tc = Database::TypeConverters::MysqlTypeConverter.new
+    @tc = Database::MysqlStrategy.new
   end
 
   context "#db_type" do

data/spec/database/migration_file_writer_spec.rb

@@ -1,16 +1,12 @@
 require "spec_helper"
 
 describe Chicago::Database::MigrationFileWriter do
-  before :each do
-    @mock_db = double(:db)
-    @mock_db.stub(:database_type).and_return(:generic)
-    @builder = described_class.new(@mock_db, "schema")
-  end
+  let(:db) { double(:db, :database_type => :generic) }
 
-  it "should return a migration file name using the timestamp of now" do
-    now = Time.local(2010, 1, 1, 12, 30)
-    Time.should_receive(:now).and_return(now)
-    @builder.migration_file.should == "schema/20100101123000_auto_migration.rb"
+  it "returns a migration file name with the current timestamp" do
+    Timecop.freeze(2010, 1, 1, 12, 30)
+    subject.migration_file("schema").
+      should == "schema/20100101123000_auto_migration.rb"
   end
 
   it "should write out a migration file generated by Sequel::MigrationBuilder" do
@@ -29,7 +25,7 @@ describe Chicago::Database::MigrationFileWriter do
     file = StringIO.new
     File.stub(:open).and_yield(file)
 
-    @builder.write_migration_file(schema)
+    subject.write_migration_file(db, schema, "directory")
 
     file.rewind
     file.read.should == "migration content"

data/spec/database/schema_generator_spec.rb

@@ -2,7 +2,7 @@ require 'spec_helper'
 require 'chicago/database/schema_generator'
 
 describe Chicago::Database::SchemaGenerator do
-  subject { described_class.new(Chicago::Database::TypeConverters::DbTypeConverter.new) }
+  subject { described_class.new(Chicago::Database::ConcreteSchemaStrategy.new) }
   
   it_behaves_like "a schema visitor"
   
@@ -41,7 +41,7 @@ describe Chicago::Database::SchemaGenerator do
     end
 
     it "should have a table type of MyISAM for mysql" do
-      subject.type_converter = Chicago::Database::TypeConverters::MysqlTypeConverter.new
+      subject.type_converter = Chicago::Database::MysqlStrategy.new
       subject.visit_fact(@fact)[:facts_sales][:table_options].should == {:engine => "myisam"}
     end
 
@@ -108,7 +108,7 @@ describe Chicago::Database::SchemaGenerator do
 
     it "should have a table type of MyISAM for mysql" do
       @dimension = @schema.define_dimension(:user)
-      subject.type_converter = Chicago::Database::TypeConverters::MysqlTypeConverter.new
+      subject.type_converter = Chicago::Database::MysqlStrategy.new
       subject.visit_dimension(@dimension)[:dimension_user][:table_options].should == {:engine => "myisam"}
     end
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: chicagowarehouse
 version: !ruby/object:Gem::Version
-  version: 0.5.1
+  version: 0.6.0
 platform: ruby
 authors:
 - Roland Swingler
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-07-03 00:00:00.000000000 Z
+date: 2014-09-01 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: sequel
@@ -30,14 +30,14 @@ dependencies:
     requirements:
     - - ! '>='
       - !ruby/object:Gem::Version
-        version: 0.3.2
+        version: 0.4.0
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - ! '>='
       - !ruby/object:Gem::Version
-        version: 0.3.2
+        version: 0.4.0
 - !ruby/object:Gem::Dependency
   name: chronic
   requirement: !ruby/object:Gem::Requirement
@@ -122,20 +122,6 @@ dependencies:
     - - ! '>='
       - !ruby/object:Gem::Version
         version: '0'
-- !ruby/object:Gem::Dependency
-  name: rcov
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ! '>='
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ! '>='
-      - !ruby/object:Gem::Version
-        version: '0'
 - !ruby/object:Gem::Dependency
   name: simplecov
   requirement: !ruby/object:Gem::Requirement
@@ -197,12 +183,13 @@ files:
 - lib/chicago/core_ext/hash.rb
 - lib/chicago/core_ext/sequel/dataset.rb
 - lib/chicago/data/month.rb
+- lib/chicago/database/concrete_schema_strategies.rb
 - lib/chicago/database/constants.rb
 - lib/chicago/database/dataset_builder.rb
 - lib/chicago/database/filter.rb
+- lib/chicago/database/index_generator.rb
 - lib/chicago/database/migration_file_writer.rb
 - lib/chicago/database/schema_generator.rb
-- lib/chicago/database/type_converters.rb
 - lib/chicago/database/value_parser.rb
 - lib/chicago/errors.rb
 - lib/chicago/query.rb
@@ -224,7 +211,7 @@ files:
 - lib/chicago/schema/table.rb
 - lib/chicago/star_schema.rb
 - spec/data/month_spec.rb
-- spec/database/db_type_converter_spec.rb
+- spec/database/concrete_schema_strategies.rb
 - spec/database/migration_file_writer_spec.rb
 - spec/database/schema_generator_spec.rb
 - spec/db_connections.yml.dist
@@ -271,4 +258,3 @@ signing_key:
 specification_version: 4
 summary: Ruby Data Warehousing
 test_files: []
-has_rdoc: 

data/lib/chicago/database/type_converters.rb

@@ -1,107 +0,0 @@
-module Chicago
-  module Database
-    module TypeConverters
-      # Generic type conversion strategy.
-      #
-      # This supplements Sequel's type conversion strategy rather than
-      # replaces it, so +:boolean+ will still return +:boolean+ rather
-      # than +tinyint(1)+ in the case of mysql.
-      class DbTypeConverter
-        # Factory method that returns an appropriate type conversion
-        # stratgey for the given database.
-        #
-        # If a database-specific strategy cannot be found, returns a
-        # generic strategy.
-        #
-        # @return [DbTypeConverter]
-        def self.for_db(db)
-          return MysqlTypeConverter.new if db.database_type == :mysql
-          self.new
-        end
-
-        # Returns a db type given a column definition
-        #
-        # @return [Symbol]
-        def db_type(column)
-          case column.column_type
-          when :integer then integer_type(column.min, column.max)
-          when :string  then string_type(column.min, column.max)
-          when :money   then :decimal
-          when :percent then :decimal
-          else
-            column.column_type
-          end
-        end
-
-        # Returns sequel table options for a dimension or fact table.
-        #
-        # None by default, but database-specific subclasses may
-        # override this.
-        #
-        # @return [Hash]
-        def table_options
-          {}
-        end
-
-        # Returns a database type for a string column.
-        #
-        # @return [Symbol]
-        def string_type(min, max)
-          min && max && min == max ? :char : :varchar 
-        end
-            
-        # Returns a database integer column type, big enough to fit
-        # values between min and max, or integer if a specific type
-        # cannot be found.
-        #
-        # @return [Symbol]
-        # @raise an ArgumentError if min or max is too large for a
-        #   single database column.
-        def integer_type(min, max)
-          signed_limit = (SMALL_INT_MAX + 1) / 2
-          if min && max && ((min >= -signed_limit && max <= signed_limit - 1) || (min >= 0 && max <= SMALL_INT_MAX))
-            :smallint
-          else
-            :integer
-          end
-        end
-      end
-
-      # MySql-specific type conversion strategy
-      class MysqlTypeConverter < DbTypeConverter
-        def db_type(column)
-          return :enum if column.elements && column.elements.size < 65_536
-          super(column)
-        end
-
-        # Returns table options for a dimension or fact table.
-        #
-        # Dimension tables are defined as MyISAM tables in MySQL.
-        def table_options
-          {:engine => "myisam"}
-        end
-
-        def integer_type(min, max)
-          return :integer unless min && max
-
-          case
-          when in_numeric_range?(min, max, TINY_INT_MAX)   then :tinyint
-          when in_numeric_range?(min, max, SMALL_INT_MAX)  then :smallint
-          when in_numeric_range?(min, max, MEDIUM_INT_MAX) then :mediumint
-          when in_numeric_range?(min, max, INT_MAX)        then :integer
-          when in_numeric_range?(min, max, BIG_INT_MAX)    then :bigint
-          else
-            raise ArgumentError.new("#{min} is too small or #{max} is too large for a single column")
-          end
-        end
-
-        private
-
-        def in_numeric_range?(min, max, unsigned_limit)
-          signed_limit = (unsigned_limit + 1) / 2
-          (min >= -signed_limit && max <= signed_limit - 1)  ||  (min >= 0 && max <= unsigned_limit)
-        end
-      end
-    end
-  end
-end