org.millstone.base.data

Provides interfaces for the MillStone data layer which contains classes for typed data values, data collections, and handlers. A {@link org.millstone.base.data.Property Property} is a simple typed data value; an {@link org.millstone.base.data.Item Item} is a collection of Properties, each corresponding to a unique identifier; a {@link org.millstone.base.data.Container Container} is a collection of identified Items with special constraints; a {@link org.millstone.base.data.Buffered Buffered} class is able to track its changes and to commit or discard them later.

Package Specification

The package contains a three-tiered structure for typed data objects and collections of them:

• The simplest of these is the {@link org.millstone.base.data.Property Property} which represents a single typed data value. A Property may be read-only in which case attempts to modify its contents will throw an exception.
• The second level of the data layer is represented by the {@link org.millstone.base.data.Item Item} which embodies a set of Properties. Each Property in an Item corresponds to a locally unique(that is, inside the Item) identifier.
• The third level is called the {@link org.millstone.base.data.Container Container} which can be visualized as a set of Items, each corresponding to a locally unique identifier. Note that the Container imposes a few restrictions on the data stored in it, see further documentation in the class specification.

In addition to these interfaces the package contains the {@link org.millstone.base.data.Buffered Buffered} interface, which defines the methods to make an object buffered, that is, track the changes to an object and allow committing or discarding them at a later time.

Provides interfaces for the MillStone validation framework. The framework defines two interfaces; one for classes that need to support external validation, and another one for the validators themselves.

Validation

The most important method defined by the {@link org.millstone.base.data.Validatable Validatable} interface is {@link org.millstone.base.data.Validatable#isValid() isValid()}. It asks all registered validators to verify if the object's value is valid or not. Note that it depends on the validators registered for a object which values are valid and which are not. For example, a null value can be valid value for one validator but invalid for another.

In addition to isValid(), Validatable defines methods to add, remove and list validators of a validatable object.

{@link org.millstone.base.data.Validator Validator} defines the interface for an external validator. These validators may be added to any Validatable object, and their task is to check, when requested, that the object which they are attached to contains a valid value. The actual validation logic is hidden in the {@link org.millstone.base.data.Validator#validate(Object) validate(Object)} method.

In addition to check(Object), Validator defines the InvalidValueException which is used to signal that a checked value is invalid, and the {@link org.millstone.base.data.Validator.Suggestive Suggestive} subinterface which includes functionality to suggest a valid value for the validated object.

Defines the interface to commit and discard changes to an object, supporting read-through and write-through modes.

Read-through mode means that the value read from the buffered object is constantly up to date with the data source. Write-through mode means that all changes to the object are immediately updated to the data source.

Since these modes are independent, their combinations may result in some behaviour that may sound surprising.

This interface defines the combination of Validatable and Buffered interfaces.

A specialized set of identified Items.

Provides a mechanism for handling a set of Properties, each associated to a locally unique identifier.

Interface for validatable objects.