Support for dumping only selected tables from database

README.markdown

@@ -24,6 +24,19 @@ Further, there are tasks db:dump and db:load which do the entire database (the e
     rake db:data:dump_dir   ->   Dump contents of database to curr_dir_name/tablename.extension (defaults to yaml)
     rake db:data:load_dir   ->   Load contents of db/data_dir into database
 
+Also, it is possible to dump only selected tables of your database:
+
+    rake db:data:dump_dir filter_tables=^table[0-5]$|other_table
+    rake db:data:dump filter_tables=^table[0-5]$|other_table
+
+(Beware that the `filter_tables` paramater is evaluated as a regular expression
+(with `Regexp.new(filter_tables)`), so:
+
+    rake db:data:dump_dir filter_tables=mytable
+
+will dump `mytable`, but also `mytable_join_other_table`.
+
+
 In addition, we have plugins whereby you can export your database to/from various formats.  We only deal with yaml and csv right now, but you can easily write tools for your own formats (such as Excel or XML).  To use another format, just load setting the "class"  parameter to the class you are using.  This defaults to "YamlDb::Helper" which is a refactoring of the old yaml_db code.  We'll shorten this to use class nicknames in a little bit.
 
 ## Examples

lib/serialization_helper.rb

@@ -3,10 +3,11 @@ module SerializationHelper
   class Base
     attr_reader :extension
 
-    def initialize(helper)
+    def initialize(helper, filter_table_names = nil)
       @dumper = helper.dumper
       @loader = helper.loader
       @extension = helper.extension
+      @dumper.filter_table_names = filter_table_names if filter_table_names
     end
 
     def dump(filename)
@@ -137,6 +138,10 @@ def self.quote_table(table)
   end
 
   class Dump
+    class << self
+      attr_accessor :filter_table_names
+    end
+
     def self.before_table(io, table)
 
     end
@@ -154,7 +159,9 @@ def self.after_table(io, table)
     end
 
     def self.tables
-      ActiveRecord::Base.connection.tables.reject { |table| ['schema_info', 'schema_migrations'].include?(table) }
+      all_tables = ActiveRecord::Base.connection.tables.reject { |table| ['schema_info', 'schema_migrations'].include?(table) }
+
+      filter_table_names ? all_tables.grep(Regexp.new(filter_table_names)) : all_tables
     end
 
     def self.dump_table(io, table)

lib/tasks/yaml_db_tasks.rake

@@ -18,14 +18,16 @@ namespace :db do
 		task :dump => :environment do
             format_class = ENV['class'] || "YamlDb::Helper"
             helper = format_class.constantize
-			SerializationHelper::Base.new(helper).dump db_dump_data_file helper.extension
+            filter = ENV['filter_tables']
+            SerializationHelper::Base.new(helper,filter).dump db_dump_data_file helper.extension
 		end
 
 		desc "Dump contents of database to curr_dir_name/tablename.extension (defaults to yaml)"
 		task :dump_dir => :environment do
             format_class = ENV['class'] || "YamlDb::Helper"
             dir = ENV['dir'] || "#{Time.now.to_s.gsub(/ /, '_')}"
-            SerializationHelper::Base.new(format_class.constantize).dump_to_dir dump_dir("/#{dir}")
+            filter = ENV['filter_tables']
+            SerializationHelper::Base.new(format_class.constantize, filter).dump_to_dir dump_dir("/#{dir}")
 		end
 
 		desc "Load contents of db/data.extension (defaults to yaml) into database"

spec/serialization_helper_base_spec.rb

@@ -33,6 +33,11 @@ def stub_helper!
          SerializationHelper::Base.new(@helper).dump_to_dir "dir_name"
       end   
 
+      it "should be able to be configured such as it will dump only selected tables" do
+        @dumper.should_receive(:"filter_table_names=").with("table_filter")
+        SerializationHelper::Base.new(@helper, "table_filter").dump_to_dir "dir_name"
+      end
+
     end
 
     context "for multi-file loads" do

spec/serialization_helper_dump_spec.rb

@@ -61,5 +61,49 @@
     end
 
 
+  context "when configured to only dump selected tables" do
+    before do
+      ActiveRecord::Base.connection.stub!(:tables).and_return([ 'mytable1', 'mytable2', 'mytable3', 'schema_info', 'schema_migrations' ])
+    end
+
+    after do
+      # Check expected tables to dump
+      SerializationHelper::Dump.tables.sort.should == @expected_tables
+
+      if @expected_tables.size > 0
+        @expected_tables.each do |table|
+          # Check actual calls to dump_table with expected tables
+          SerializationHelper::Dump.should_receive(:dump_table).with(nil, table)
+        end
+      else
+        SerializationHelper::Dump.should_not_receive(:dump_table).with(anything)
+      end
 
+      SerializationHelper::Dump.dump(nil)
+
+      # Restore default behaviour
+      SerializationHelper::Dump.filter_table_names = nil
+    end
+
+    it "should dump every table if filter not set" do
+      SerializationHelper::Dump.filter_table_names = nil
+      @expected_tables =  ['mytable1', 'mytable2', 'mytable3']
+    end
+
+    it "should dump matching tables if filter is set as a matching string" do
+      SerializationHelper::Dump.filter_table_names = "mytable"
+      @expected_tables =  ['mytable1', 'mytable2', 'mytable3']
+    end
+
+    it "should dump matching tables if filter is set as a regular expression string" do
+      SerializationHelper::Dump.filter_table_names = "mytable[1-2]"
+      @expected_tables =  ['mytable1', 'mytable2']
+    end
+
+    it "should not dump any table if filter does not match any table" do
+      SerializationHelper::Dump.filter_table_names = "wadustable"
+      @expected_tables =  []
+    end
+
+  end
 end