1. Introduction
Passing immutable data between objects is one of the most common, but mundane tasks in many Java applications.
Prior to Java 14, this required the creation of a class with boilerplate fields and methods, which were susceptible to trivial mistakes and muddled intentions.
With the release of Java 14, we can now use records to remedy these problems.
In this tutorial, we'll look at the fundamentals of records, including their purpose, generated methods, and customization techniques.
2. Purpose
Commonly, we write classes to simply hold data, such as database results, query results, or information from a service.
In many cases, this data is immutable, since immutability ensures the validity of the data without synchronization.
To accomplish this, we create data classes with the following:
- private, final field for each piece of data
- getter for each field
- public constructor with a corresponding argument for each field
- equals method that returns true for objects of the same class when all fields match
- hashCode method that returns the same value when all fields match
- toString method that includes the name of the class and the name of each field and its corresponding value
For example, we can create a simple Person data class, with a name and an address:
public class Person { private final String name; private final String address; public Person(String name, String address) { this.name = name; this.address = address; } @Override public int hashCode() { return Objects.hash(name, address); } @Override public boolean equals(Object obj) { if (this == obj) { return true; } else if (!(obj instanceof Person)) { return false; } else { Person other = (Person) obj; return Objects.equals(name, other.name) && Objects.equals(address, other.address); } } @Override public String toString() { return "Person [name=" + name + ", address=" + address + "]"; } // standard getters }
While this accomplishes our goal, there are two problems with it:
- There is a lot of boilerplate code
- We obscure the purpose of our class – to represent a person with a name and address
In the first case, we have to repeat the same tedious process for each data class, monotonously creating a new field for each piece of data, creating equals, hashCode, and toString methods, and creating a constructor that accepts each field.
While IDEs can automatically generate many of these classes, they fail to automatically update our classes when we add a new field. For example, if we add a new field, we have to update our equals method to incorporate this field.
In the second case, the extra code obscures that our class is simply a data class that has two String fields: name and address.
A better approach would be to explicitly declare that our class is a data class.
3. The Basics
As of JDK 14, we can replace our repetitious data classes with records. Records are immutable data classes that require only the type and name of fields.
The equals, hashCode, and toString methods, as well as the private, final fields, and public constructor, are generated by the Java compiler.
To create a Person record, we use the record keyword:
public record Person (String name, String address) {}
3.1. Constructor
Using records, a public constructor – with an argument for each field – is generated for us.
In the case of our Person record, the equivalent constructor is:
public Person(String name, String address) { this.name = name; this.address = address; }
This constructor can be used in the same way as a class to instantiate objects from the record:
Person person = new Person("John Doe", "100 Linda Ln.");
3.2. Getters
We also receive public getters methods – whose names match the name of our field – for free.
In our Person record, this means a name() and address() getter:
@Test public void givenValidNameAndAddress_whenGetNameAndAddress_thenExpectedValuesReturned() { String name = "John Doe"; String address = "100 Linda Ln."; Person person = new Person(name, address); assertEquals(name, person.name()); assertEquals(address, person.address()); }
3.3. equals
Additionally, an equals method is generated for us.
This method returns true if the supplied object is of the same type and the values of all of its fields match:
@Test public void givenSameNameAndAddress_whenEquals_thenPersonsEqual() { String name = "John Doe"; String address = "100 Linda Ln."; Person person1 = new Person(name, address); Person person2 = new Person(name, address); assertTrue(person1.equals(person2)); }
If any of the fields differ between two Person instances, the equals method will return false.
3.4. hashCode
Similar to our equals method, a corresponding hashCode method is also generated for us.
Our hashCode method returns the same value for two Person objects if all of the field values for both object match (barring collisions due to the birthday paradox):
@Test public void givenSameNameAndAddress_whenHashCode_thenPersonsEqual() { String name = "John Doe"; String address = "100 Linda Ln."; Person person1 = new Person(name, address); Person person2 = new Person(name, address); assertEquals(person1.hashCode(), person2.hashCode()); }
The hashCode value will differ if any of the field values differ.
3.5. toString
Lastly, we also receive a toString method that results in a string containing the name of the record, followed by the name of each field and its corresponding value in square brackets.
Therefore, instantiating a Person with a name of “John Doe” and an address of “100 Linda Ln.” results in the following toString result:
Person[name=John Doe, address=100 Linda Ln.]
4. Constructors
While a public constructor is generated for us, we can still customize our constructor implementation.
This customization is intended to be used for validation and should be kept as simple as possible.
For example, we can ensure that the name and address provided to our Person record are not null using the following constructor implementation:
public record Person(String name, String address) { public Person { Objects.requireNonNull(name); Objects.requireNonNull(address); } }
We can also create new constructors with different arguments by supplying a different argument list:
public record Person(String name, String address) { public Person(String name) { this(name, "Unknown"); } }
As with class constructors, the fields can be referenced using the this keyword (for example, this.name and this.address) and the arguments match the name of the fields (that is, name and address).
Note that creating a constructor with the same arguments as the generated public constructor is valid, but this requires that each field be manually initialized:
public record Person(String name, String address) { public Person(String name, String address) { this.name = name; this.address = address; } }
Additionally, declaring a no-argument constructor and one with an argument list matching the generated constructor results in a compilation error.
Therefore, the following will not compile:
public record Person(String name, String address) { public Person { Objects.requireNonNull(name); Objects.requireNonNull(address); } public Person(String name, String address) { this.name = name; this.address = address; } }
5. Static Variables & Methods
As with regular Java classes, we can also include static variables and methods in our records.
We declare static variables using the same syntax as a class:
public record Person(String name, String address) { public static String UNKNOWN_ADDRESS = "Unknown"; }
Likewise, we declare static methods using the same syntax as a class:
public record Person(String name, String address) { public static Person unnamed(String address) { return new Person("Unnamed", address); } }
We can then reference both static variables and static methods using the name of the record:
Person.UNKNOWN_ADDRESS Person.unnamed("100 Linda Ln.");
6. Conclusion
In this article, we looked at the record keyword introduced in Java 14, including their fundamental concepts and intricacies.
Using records – with their compiler-generated methods – we can reduce boilerplate code and improve the reliability of our immutable classes.
The code and examples for this tutorial can be found over on GitHub.