Chapter 4: Creating Flexible Interfaces

Question 1

What is an interface?

Answer 1

An interface is the means objects use to talk. Design deals not only with what objects an object knows (its responsibilities) and who the object knows (its dependencies) but also how it talks to another.

Question 2

Mention two concepts on how we can uses an interface

Answer 2

It can refer to:

• The methods within a class and their public or private exposition
• The one that spans across classes and that is independent of any single class. In this sense an interface is a set of methods that define it.

Question 3

What are the features of the public interface of a class?

Answer 3

• Reveal its primary responsibility
• Are expected to be invoked by others
• Will not change in a whin
• Are safe for others to depend on
• Are thoroughly documented in the tests

Question 4

What are the features of the private interface of a class?

Answer 4

• Handle implementation details
• Are not expected to be sent by other objects
• Can change for any reason whatsoever
• Are unsafe for others to depend on
• May not even be referenced in the tests

Question 5

What is a public interface within the class?

Answer 5

It is a contract that articulates the resposibilities of the class.

Question 6

What is the relationship between Single Responsibility approach and the public interface?

Answer 6

Methods in the public interface of a class should read like a description of responsibilities.

Question 7

What is the relationship between Managing Dependencies and the class’ interface?

Answer 7

The idea of depends on classes that change less than yours do can be applied to the methods in a class. Methods in the public interface are stable and private ones have more likelihood of change.

Question 8

What are Domain Objects?

Answer 8

They are classes that are easily identified in the requirements because they are nouns in the application that have both data and behavior. They stand for big, visible real-world things.

Question 9

Why message-based perspective is better than class-based ones?

Answer 9

Message-based perspective yield more flexible applications than class-based perspective.

Question 10

Why the sequence diagram is suitable for the message-based design?

Answer 10

The sequence diagram allows checking the interaction between objects and the messages passed between them.

Question 11

Why is it important to ask for “What” instead of telling “How”?

Answer 11

The “What” allows to create reusable classes with well-defined public interfaces.

Question 12

What is the context of a class?

Answer 12

The things a class know about other objects make up its context.

Question 13

Why does a context is important?

Answer 13

The context that an object expect has a direct effect on how difficult it is to reuse. Object with a simple context are easy to use and test.

Question 14

What does seeking context independence means?

Answer 14

It means creating objects with a simple context where they aren’t holding onto other objects.

Question 15

What does trusting other objects means?

Answer 15

It means thinking about what instead of how. When we trust another object will do its work, we can just ask (the what) the things we want and the context will be much simplier.

Question 16

What is the purpose of a sequence diagram?

Answer 16

Act as the vehicle for exposing, experiment with, and ultimately defining the public interfaces.

Question 17

What is the focus of the message-based approach?

Answer 17

It is to answer the question: “I need to send this message, who should respond to it?”

Question 18

What are Private, Protected and Public key words?

Answer 18

They are access modifiers

Question 19

How does the Private keyword work?

Answer 19

It denotes the least stable kind of method and provides the most restricted visibility.
Private methods must be called with an implicit receiver (that is ‘self’) or, inversely, may never be called with an explicit receiver.

Question 20

In what cases a private method can be sent?

Answer 20

If we have a private method called fun_factor in the Trip class, we can send:
- fun_factor, defaulting to self (the implicit receiver) from within instances of Trip and its subclasses
But we may not send:
- self.fun_factor from within Trip
- a_trip.fun_factor from another object

Question 21

How does the Protected keyword work?

Answer 21

It also indicates an unstable method like private ones but with slightly different visibility restrictions. Protected methods allow explicit receivers as long as the receiver is self or an instance of the same class or subclass of self

Question 22

In what cases a protected method can be sent?

Answer 22

If we have a protected method called fun_factor in the Trip class, we can send:

• self.fun_factor
• a_trip.fun_factor, but only from within a class where self is the same kind of thing (class or subclass) as a_trip

Question 23

What do you do if a class doesn’t offert a public interface or an ill-defined one?

Answer 23

We must strive for minimize the context our interfaces require from others. We can always create our own public interfaces if other classes don’t provide one or they are ill-defined.

Question 24

What is the Law of Demeter?

Answer 24

It is a set of coding rules that results in loosely coupled objects.

Question 25

When are we violating Demeter?

Answer 25

Demeter restricts the set of objects to which a method may send messages; it prohibits routing a message to a third object via a second one of a different type.
When we have a long chaining list of methods (know as train wrecks) we might be violating Demeter.

Question 26

How do we avoid the train wrecks Demeter violation?

Answer 26

We can use Delegation but beware, using delegation to hide tight coupling is not the same as decopling de code.

Question 27

Why does a train wrecks represent a Demeter violation?

Answer 27

Because we are are trying to get a distant behavior through the chain of objects.