This is a DataMapper plugin that provides validations for DataMapper model classes.
DataMapper validation capabilities are automatically available for DataMapper resources when you require dm-validations' in your application. There is no need to manually include anything, every DataMapper::Resource will be able to handle validations once this gem got required.
Specifying Model Validations
There are two primary ways to implement validations for your models
1) Placing validation methods with properties as params in your class
require 'dm-core' require 'dm-validations' class ProgrammingLanguage include ::Resource property :name, String validates_presence_of :name end
2) Using auto-validations, please see DataMapper::Validation::AutoValidations. Note that not all validations that are provided via validation methods, are also available as autovalidation options. If they are available, they're functionally equivalent though.
class ProgrammingLanguage include ::Resource property :name, String, :required => true end
See data_mapper/validation/macros.rb for to learn about the complete collection of validation rules available.
DataMapper validations, when included, alter the default save/create/update process for a model. Unless you specify a context the resource must be valid in the :default context before saving.
You may manually validate a resource using the valid? method, which will return true if the resource is valid, and false if it is invalid.
Working with Validation Errors
If your validators find errors in your model, they will populate the DataMapper::Validation::ViolationSet object that is available through each of your models via calls to your model's errors method.
my_account = Account.new(:name => "Jose") if my_account.save # my_account is valid and has been saved else my_account.errors.each do |e| puts e end end
See DataMapper::Validation::ViolationSet for all you can do with your model's errors method.
DataMapper Validation also provide a means of grouping your validations into contexts. This enables you to run different sets of validations when you need it. For instance, the same model may not only behave differently when initially saved or saved on update, but also require special validation sets for publishing, exporting, importing and so on.
Again, using our example for pure Ruby class validations:
class ProgrammingLanguage include ::Resource property :name, String def ensure_allows_manual_memory_management # ... end def ensure_allows_optional_parentheses # ... end validates_presence_of :name validates_with_method :ensure_allows_optional_parentheses, :when => [:implementing_a_dsl] validates_with_method :ensure_allows_manual_memory_management, :when => [:doing_system_programming] end
ProgrammingLanguage instance now use #valid? method with one of two context symbols:
@ruby.valid?(:implementing_a_dsl) # => true @ruby.valid?(:doing_system_programming) # => false @c.valid?(:implementing_a_dsl) # => false @c.valid?(:doing_system_programming) # => true
Each context causes different set of validations to be triggered. If you don't specify a context using :when, :on or :group options (they are all aliases and do the same thing), default context name is :default. When you do model.valid? (without specifying context explicitly), again, :default context is used. One validation can be used in two, three or five contexts if you like:
class Book include ::::Resource property :id, Serial property :name, String property :agreed_title, String property :finished_toc, Boolean # used in all contexts, including default validates_presence_of :name, :when => [:default, :sending_to_print] validates_presence_of :agreed_title, :when => [:sending_to_print] validates_with_block :toc, :when => [:sending_to_print] do if self.finished_toc [true] else [false, "TOC must be finalized before you send a book to print"] end end end
In the example above, name is validated for presence in both :default context and :sending_to_print context, while TOC related block validation and title presence validation only take place in :sending_to_print context.