We are going to dive deep into hashCode method in Java.
How hashCode influences performance
Object class contains default implementation of hasCode method. We can override that method and implement.
class Test {
@Override
public int hashCode() {
return 1;
}
}
Test test = new Test();
System.out.println(test.hashCode());
Here is the output.
1
There is a way to get original hashCode of the object before we overridden it. We can use indentityHashCode method from System class.
int originalHashCode = System.identityHashCode(test);
System.out.println()
We can verify that identityHashCode method really prints the original hash code. Just get the hashCode of parent in the hasCode method.
class Test {
@Override
public int hashCode() {
System.out.println(super.hashCode())
return 1;
}
}
If we provide weak implemention of hashCode method, we might degrade peformance of hash map. Lets try to create many instances of Test class and insert them into HashMap. This code takes about 1 second to insert 10 000 items into hash map. All because of wrong implementation of hashCode.
class Test {
@Override
public int hashCode() {
return 1;
}
}
long start = new Date().getTime();
Set map = new HashSet();
for (int i = 0; i < 10000; i++) {
map.add(new Test());
}
System.out.println(new Date().getTime() - start);
// prints out 915
Here is the code that leaves implementation of hashCode to Object class. This one will insert 10 000 items into hash map in about 5 milliseconds.
class Test {
}
long start = new Date().getTime();
Set map = new HashSet();
for (int i = 0; i < 10000; i++) {
map.add(new Test());
}
System.out.println(new Date().getTime() - start);
// prints out 5
What should be the proper implementation of hashCode for Test? For that class, we shouldn't touch the hashCode method.
In order to override hashCode, we need to get better example. Lets create User class that will contain couple of fields. We want to use those fields to calculate hashCode. We can Objects.hash method since Java 7.
import java.util.Objects;
public class User {
private String name;
private int age;
private String passport;
@Override
public int hashCode() {
return Objects.hash(name, age, passport);
}
}
Lets create the same performance test we did with Test class.
public class User {
String name;
int age;
String passport;
@Override
public int hashCode() {
return Objects.hash(name, age, passport);
}
}
long start = new Date().getTime();
Set map = new HashSet();
for (int i = 0; i < 10000; i++) {
User user = new User();
user.name = String.valueOf(i);
user.age = i;
user.passport = String.valueOf(i);
map.add(user);
}
System.out.println(new Date().getTime() - start);
// prints out 33
Now the hashCode method is implemented correctly and we do not have issue performance issue when inserting values into hash set.
Lets try to verify if we are able to find out if the user is present in that hash set.
User user = new User();
user.name = "9999";
user.age = 9999;
user.passport = "9999";
boolean contains = map.contains(user);
System.out.println(contains);
// prints out false
It says that the user is not present in the hash set, because we didn't overridden equals method. Lets do it now.
public class User {
String name;
int age;
String passport;
@Override
public boolean equals(Object o) {
if (o == this) return true;
if (!(o instanceof User)) {
return false;
}
User user = (User) o;
return age == o.age &&
Objects.equals(name, user.name) &&
Objects.equals(passport, user.passport);
}
@Override
public int hashCode() {
return Objects.hash(name, age, passport);
}
}
Set<User> map = new HashSet<>();
for (int i = 0; i < 10000; i++) {
User user = new User();
user.name = String.valueOf(i);
user.age = i;
user.passport = String.valueOf(i);
map.add(user);
}
User user = new User();
user.name = "9999";
user.age = 9999;
user.passport = "9999";
boolean contains = map.contains(user);
System.out.println(contains);
// prints out true