rod 0.6.2 → 0.6.3

data/.travis.yml

@@ -0,0 +1 @@
+rvm: 1.9.2

data/README.rdoc

@@ -18,7 +18,7 @@ fast access for data, which rarely changes.
 * compatibility check of data model
 * autogeneration of model (based on the database metadata)
 * automatic model migrations (addition/removal of properties so far)
-* append of the database (new objects, new elements in plural associations)
+* full update of the database (removal of objects not available yet)
 * databases interlinking (via direct links or inverted indices)
 
 == PROBLEMS
@@ -39,9 +39,9 @@ The primary reason for designing it was to create storage facility for
 natural language dictionaries and corpora. The data in a fully fledged dictionary
 is interconnected in many ways, thus the relational model (joins) introduces
 unacceptable performance hit. The size of corpora forces them to be kept
-on disks. The in-memory data bases are unacceptable for larg corpora and
-would require the data to be kept mostly in the operational memory,
-which is not needed, while accessing dictionaries (in most cases only a fraction
+on disks. The in-memory data bases are unacceptable for larg corpora.
+They would also require the data of a dictionary to be kept mostly in the
+operational memory, which is not needed (in most cases only a fraction
 of the data is used at the same time). That's why a storage facility which minimizes the
 number of disk reads was designed. The Ruby interface facilitates it's usage.
 
@@ -51,6 +51,7 @@ number of disk reads was designed. The Ruby interface facilitates it's usage.
 * RubyInline
 * english
 * ActiveModel
+* bsearch
 
 == INSTALL
 
@@ -69,7 +70,7 @@ Grab from rubygems:
 
  class User < Model
    field :name, :string
-   field :surname, :string, :index => true
+   field :surname, :string, :index => :flat
    field :age, :integer
    has_one :account
    has_many :files
@@ -77,12 +78,12 @@ Grab from rubygems:
 
  class Account < Model
    field :email, :string
-   field :login, :string, :index => true
+   field :login, :string, :index => :flat
    field :password, :string
  end
 
  class File < Model
-   field :title, :string, :index => true
+   field :title, :string, :index => :flat
    field :data, :string
  end
 
@@ -113,7 +114,7 @@ Grab from rubygems:
    puts "Name: #{user.name} surname: #{user.surname}"
    puts "login: #{user.account.login} e-mail: #{user.account.email}"
    user.files.each do |file|
-     puts "File: #{file.title}
+     puts "File: #{file.title}"
    end
  end
 
@@ -130,7 +131,7 @@ You'll need bundler installed:
 
 Then you have to fetch all the dependencies:
 
-  bunlde
+  bundle
 
 To run all the test simple type rake:
 

data/Rakefile

@@ -47,12 +47,17 @@ task :test do
   sh "ruby tests/generate_classes_rewrite.rb"
   sh "ruby tests/generate_classes_rewrite.rb"
   sh "ruby tests/generate_classes_verify.rb"
-  sh "ruby tests/migration_create.rb"
-  sh "ruby tests/migration_migrate.rb"
-  sh "ruby tests/migration_verify.rb"
+  sh "ruby tests/migration_create.rb 1000"
+  sh "ruby tests/migration_migrate.rb 1000"
+  sh "ruby tests/migration_verify.rb 1000"
+  sh "ruby tests/missing_class_create.rb"
+  sh "ruby tests/missing_class_verify.rb"
+  sh "ruby tests/properties_order_create.rb"
+  sh "ruby tests/properties_order_verify.rb"
   sh "ruby tests/unit/model.rb"
   sh "ruby tests/unit/model_tests.rb"
   sh "ruby tests/unit/database.rb"
+  sh "ruby tests/unit/abstract_database.rb"
 end
 
 # Should be removed some time -- specs should cover all these cases
@@ -61,17 +66,22 @@ task :regression_test do
   sh "ruby tests/check_strings.rb"
 end
 
+desc "Run all specs without the ignored ones"
 task :spec do
   sh "bundle exec cucumber --tags ~@ignore features/*"
 end
 
-# Work in progress
+desc "Run only work-in-progress specs"
 task :wip do
   sh "bundle exec cucumber --tags @wip features/*"
 end
 
-desc "Clean"
+desc "Clean all gems"
 task :clean do
   sh "rm #$gem_name*.gem" 
 end
 
+desc "Show changelog from the last release"
+task :changelog do
+  sh "git log v#{Rod::VERSION}.. --pretty=%s | tee"
+end

data/changelog.txt

@@ -1,3 +1,21 @@
+0.6.3
+- #94 indexing of unstored object in plural assoc
+- #128 incompatible schema message is more specific
+- #129 check properties order
+- #131 check for missing classes
+- #72 update of objects' state
+- #117 get rid of signed/unsigned comparison warning
+- #135 refactoring of indexes
+- #140 remove key argument from AbstractDatabase#read_index
+- #149 CollectionProxy insert, delete, delete_at
+- #154 fix configuration of mocha for cucumber
+- #153 fix: field values doesn't match
+- #153 111122223333 is too big for ulong on 32bit systems
+- #18 use latest RubyInline and fix 'require on ... failed'
+- #152 travis-ci configuration: ruby 1.9.2
+- #151 missing rake dependency
+- #130 fix: canonicalize path
+- #124 implement empty? for CollectionProxy
 0.6.2
 - #127 default implementation of inspect
 - #123 fix: warning on has_rdoc in gemspec

data/features/append.feature

@@ -183,8 +183,6 @@ Feature: database with append
     And there should be 1 Caveman with the first Automobile as automobiles
     And there should be 1 Caveman with the second Automobile as automobiles
 
-  # Enable for #94
-  @ignore
   Scenario: append of has many associations with indexing with an unstored object
       Same as above, but with an object which is appended to the collection
       while it is not yet stored in the DB.

data/features/basic.feature

@@ -45,7 +45,7 @@ Feature: Store and load small amount of data from one class
     And I create a Caveman
     And his name is 'Fred'
     And his age is '25'
-    And his identifier is '111122223333'
+    And his identifier is '111222333'
     And his height is '1.86'
     And his symbol is ':fred'
     # nil is converted to an empty string, consider using object field
@@ -59,7 +59,7 @@ Feature: Store and load small amount of data from one class
     Then there should be 1 Caveman
     And the name of the first Caveman should be 'Fred'
     And the age of the first Caveman should be '25'
-    And the identifier of the first Caveman should be '111122223333'
+    And the identifier of the first Caveman should be '111222333'
     And the height of the first Caveman should be '1.86'
     And the symbol of the first Caveman should be ':fred'
     And the empty_string of the first Caveman should be ''
@@ -100,11 +100,11 @@ Feature: Store and load small amount of data from one class
     And I create a Caveman
     And his name is 'Fred'
     And his age is '25'
-    And his identifier is '111122223333'
+    And his identifier is '111222333'
     And his height is '1.86'
     Then his name should be 'Fred'
     And his age should be '25'
-    And his identifier should be '111122223333'
+    And his identifier should be '111222333'
     And his height should be '1.86'
 
     When database is created
@@ -134,9 +134,9 @@ Feature: Store and load small amount of data from one class
     And I create a Caveman
     And I fetch the first Caveman created
     Then his name should be ''
-    And his age should be nil
-    And his identifier should be nil
-    And his height should be nil
+    And his age should be '0'
+    And his identifier should be '0'
+    And his height should be '0'
     And his symbol should be nil
     And his item should be nil
     And his items should be empty

data/features/collection_proxy.feature

@@ -0,0 +1,140 @@
+Feature: collection proxy specification
+
+  Scenario: appending items
+    Given the initial size of the collection proxy is 10
+    When I append a new item 10 times
+    Then the size of the collection proxy should be 20
+    And the collection proxy should behave like an array
+
+    When I append a new item 10 times
+    Then the size of the collection proxy should be 30
+    And the collection proxy should behave like an array
+
+  Scenario: inserting items
+    Given the initial size of the collection proxy is 10
+    When I insert a new item at position 0
+    Then the size of the collection proxy should be 11
+    And the collection proxy should behave like an array
+    When I insert a new item at position 11
+    Then the size of the collection proxy should be 12
+    And the collection proxy should behave like an array
+    When I insert a new item at position 5
+    Then the size of the collection proxy should be 13
+    And the collection proxy should behave like an array
+    When I insert an item with rod_id = 1 at position 5 3 times
+    Then the size of the collection proxy should be 16
+    And the collection proxy should behave like an array
+
+  Scenario: deleting items
+    Given the initial size of the collection proxy is 5
+    When I delete an item at position 0 2 times
+    Then the size of the collection proxy should be 3
+    And the collection proxy should behave like an array
+    When I delete an item with rod_id = 3
+    Then the size of the collection proxy should be 2
+    And the collection proxy should behave like an array
+    When I delete an item with rod_id = 3
+    Then the size of the collection proxy should be 2
+    And the collection proxy should behave like an array
+    When I delete an item with rod_id = 4
+    Then the size of the collection proxy should be 1
+    And the collection proxy should behave like an array
+    When I delete an item at position 1
+    Then the size of the collection proxy should be 1
+    And the collection proxy should behave like an array
+    When I delete an item at position 0
+    Then the collection proxy should be empty
+    And the collection proxy should behave like an array
+
+  Scenario: deleting and inserting items
+    Given the initial size of the collection proxy is 5
+    When I delete an item with rod_id = 1
+    Then the size of the collection proxy should be 4
+    And the collection proxy should behave like an array
+    When I insert a new item at position 0
+    Then the size of the collection proxy should be 5
+    And the collection proxy should behave like an array
+    When I insert an item with rod_id = 6 at position 1
+    Then the size of the collection proxy should be 6
+    And the collection proxy should behave like an array
+    When I delete an item at position 2
+    Then the size of the collection proxy should be 5
+    And the collection proxy should behave like an array
+    When I insert a new item at position 2
+    Then the size of the collection proxy should be 6
+    And the collection proxy should behave like an array
+    When I delete an item with rod_id = 6
+    Then the size of the collection proxy should be 4
+    And the collection proxy should behave like an array
+    When I delete an item at position 1
+    Then the size of the collection proxy should be 3
+    And the collection proxy should behave like an array
+    When I delete an item at position 0
+    Then the size of the collection proxy should be 2
+    And the collection proxy should behave like an array
+    When I delete an item at position 0
+    Then the size of the collection proxy should be 1
+    And the collection proxy should behave like an array
+    When I delete an item at position 0
+    Then the size of the collection proxy should be 0
+    And the collection proxy should behave like an array
+    When I insert a new item at position 0
+    Then the size of the collection proxy should be 1
+    And the collection proxy should behave like an array
+
+  Scenario: deleting and appending items
+    Given the initial size of the collection proxy is 5
+    When I append a new item 5 times
+    Then the size of the collection proxy should be 10
+    And the collection proxy should behave like an array
+    When I delete an item at position 0
+    Then the size of the collection proxy should be 9
+    And the collection proxy should behave like an array
+    When I delete an item with rod_id = 6
+    Then the size of the collection proxy should be 8
+    And the collection proxy should behave like an array
+    When I append a new item
+    Then the size of the collection proxy should be 9
+    And the collection proxy should behave like an array
+
+  Scenario: inserting, deleting and appending items
+    Given the initial size of the collection proxy is 5
+    When I append a new item 5 times
+    Then the size of the collection proxy should be 10
+    And the collection proxy should behave like an array
+    When I insert a new item at position 2 3 times
+    Then the size of the collection proxy should be 13
+    And the collection proxy should behave like an array
+    When I insert an item with rod_id = 6 at position 12
+    Then the size of the collection proxy should be 14
+    And the collection proxy should behave like an array
+    When I insert a new item at position 0
+    Then the size of the collection proxy should be 15
+    And the collection proxy should behave like an array
+    When I delete an item with rod_id = 1
+    Then the size of the collection proxy should be 14
+    And the collection proxy should behave like an array
+    When I delete an item with rod_id = 6
+    Then the size of the collection proxy should be 12
+    And the collection proxy should behave like an array
+    When I delete an item with rod_id = 6
+    Then the size of the collection proxy should be 12
+    And the collection proxy should behave like an array
+    When I delete an item at position 5 2 times
+    Then the size of the collection proxy should be 10
+    And the collection proxy should behave like an array
+    When I insert a new item at position 0 5 times
+    Then the size of the collection proxy should be 15
+    And the collection proxy should behave like an array
+    When I delete an item at position 10 5 times
+    Then the size of the collection proxy should be 10
+    And the collection proxy should behave like an array
+    When I append a new item 5 times
+    Then the size of the collection proxy should be 15
+    And the collection proxy should behave like an array
+    When I delete an item at position 5 15 times
+    Then the size of the collection proxy should be 5
+    And the collection proxy should behave like an array
+    When I delete an item at position 0 5 times
+    Then the collection proxy should be empty
+    And the collection proxy should behave like an array

data/features/flat_indexing.feature

@@ -17,19 +17,19 @@ Feature: Access to objects with indexed fields
     And I create a Caveman
     And his name is 'Fred'
     And his age is '25'
-    And his identifier is '111122223333'
+    And his identifier is '111222333'
     And his height is '1.86'
     And I store him in the database
     And I create another Caveman
     And his name is 'Barney'
     And his age is '26'
-    And his identifier is '111122224444'
+    And his identifier is '111222444'
     And his height is '1.67'
     And I store him in the database
     And I create another Caveman
     And his name is 'Wilma'
     And his age is '25'
-    And his identifier is '111122225555'
+    And his identifier is '111222555'
     And his height is '1.67'
     And I store him in the database
     And I reopen database for reading
@@ -39,9 +39,9 @@ Feature: Access to objects with indexed fields
     And there should be 1 Caveman with 'Barney' name
     And there should be 2 Caveman(s) with '25' age
     And there should be 1 Caveman with '26' age
-    And there should be 1 Caveman with '111122223333' identifier
-    And there should be 1 Caveman with '111122224444' identifier
-    And there should be 1 Caveman with '111122225555' identifier
+    And there should be 1 Caveman with '111222333' identifier
+    And there should be 1 Caveman with '111222444' identifier
+    And there should be 1 Caveman with '111222555' identifier
     And there should be 2 Caveman(s) with '1.67' height
     And there should be 1 Caveman with '1.86' height
 
@@ -92,13 +92,14 @@ Feature: Access to objects with indexed fields
       | Adam | Parker  | adam  | 12   |
       | Adam |         | noob1 | 33   |
       |      |         | noob2 | -1   |
+      |      | Adam    | noob1 | 33   |
     And I reopen database for reading
-    Then there should be 5 Caveman(s)
+    Then there should be 6 Caveman(s)
     And there should be 1 Caveman with 'John' name
     And there should be 2 Caveman(s) with 'Adam' name
     And there should be 2 Caveman(s) with '12' age
     And there should be 1 Caveman with '-1' age
-    And there should be 1 Caveman with '' name
+    And there should be 2 Caveman with '' name
     And there should be 2 Caveman(s) with '' surname
 
   Scenario: multiple object with indexed fields

data/features/{fred.feature → persistence.feature}

@@ -1,7 +1,5 @@
-Feature: ROD Storage
-  In order to persist data
-  Potential users
-  Must be able to store and load their objects.
+Feature: Persistence model
+    This feature defines the persistence model of ROD.
   Background:
     Given the library works in development mode
     And the class space is cleared
@@ -9,10 +7,9 @@ Feature: ROD Storage
     And a class Fred has an age field of type integer
     And a class Fred has a sex field of type string with flat index
 
-  Scenario Outline: Store Fred and modify his age
-      So far Rod doesn't support changing of objects (especially
-      plural associations). To keep the behavior consisten for
-      all types of data, fields shouldn't be modifiable as well.
+  Scenario Outline: Persistence of unstored changes
+      If there are any changes made to the object after it has
+      been persisted, they are lost when the database is closed.
     When database is created
     And I create a Fred
     And his age is '<init_age>'

data/features/{assoc_indexing.feature → relationship_indexing.feature}

@@ -64,3 +64,39 @@ Feature: Access to objects with indexed associations
     And there should exist a Caveman with the second Automobile as automobiles
     And there should be 1 Caveman with the first Automobile as automobiles
     And there should be 1 Caveman with the second Automobile as automobiles
+
+  Scenario: indexing of plural associations with fresh objects
+    Given the class space is cleared
+    And the model is connected with the default database
+    And a class Automobile has a name field of type string
+    And a class Caveman has a name field of type string
+    And a class Caveman has many automobiles with flat index
+    When database is created
+    And I create an Automobile
+    And its name is 'Prehistoric car'
+    And I create another Automobile
+    And its name is 'Modern car'
+    And I create a Caveman
+    And his name is 'Fred'
+    And his automobiles contain the first Automobile created
+    And his automobiles contain the second Automobile created
+    And I store him in the database
+    And I create another Caveman
+    And his name is 'Allen'
+    And his automobiles contain the second Automobile created
+    And I store him in the database
+    And I fetch the first Automobile created
+    And I store it in the database
+    And I fetch the second Automobile created
+    And I store it in the database
+    And I create another Caveman
+    And her name is 'Linda'
+    And her automobiles contain the first Automobile created
+    And I store her in the database
+    And I reopen database for reading
+    Then there should be 3 Caveman(s)
+    And there should be 2 Automobile(s)
+    And there should exist a Caveman with the first Automobile as automobiles
+    And there should exist a Caveman with the second Automobile as automobiles
+    And there should be 2 Caveman(s) with the first Automobile as automobiles
+    And there should be 2 Caveman(s) with the second Automobile as automobiles