Table of Contents
Java 12 introduced a new feature into the Java language called switch expressions, which can be used to simplify many switch statements. Switch statements are often used to avoid long chains of if
and else if
statements, generally making your code more readable.
That being said, switch statements can be verbose in their own way, and the strict requirements for placing break
statements often make them error-prone. Switch expressions were designed to offer a more concise and less error-prone alternative to switch statements.
Switch statements vs switch expressions
The main difference between a switch expression and a switch statement is that while a switch statement can be used to update the value of a predefined variable, a switch expression is assigned to a variable. This is possible because a switch expression evaluates to a specific value. Additionally, switch expressions introduced a new arrow syntax that condenses the code, makes it more readable, and eliminates the need for break
statements. Let’s look at an example that demonstrates these differences.
We begin with an enumeration of various things that are going to be taste-tested. Our switch statement and switch expressions are going to assign a taste rating as an integer from 1 to 10, with 1 being utterly disgusting and 10 being absolutely delicious. First, we’ll look at how this would commonly be written as a switch statement:
private enum ThingsToTaste {PIZZA, BROCCOLI, STEAK, SUGAR, DIRT, MEATBALLS, CHOCOLATE}
int tasteValue = 0;
ThingsToTaste taste = ThingsToTaste.DIRT;
switch (taste) {
case SUGAR:
case PIZZA:
case CHOCOLATE:
tasteValue = 10;
break;
case MEATBALLS:
case STEAK:
tasteValue = 7;
break;
case BROCCOLI:
tasteValue = 4;
break;
case DIRT:
tasteValue = 1;
break;
default:
throw new IllegalStateException("Invalid tastable object: " + taste);
}
System.out.println(taste + ": " + tasteValue);
Now let’s contrast this with a switch expression.
int tasteValue = switch (taste) {
case SUGAR, PIZZA, CHOCOLATE -> 10;
case MEATBALLS, STEAK -> 7;
case BROCCOLI -> 4;
case DIRT -> 1;
default -> throw new IllegalStateException("Invalid tastable object: " + taste);
};
As you can see, this is way shorter. Let’s go through what changed! First, the tasteValue
variable did not have to be initialized before the switch. Instead, the entire switch expression is assigned to be the value of tasteValue
. This works because the switch expression will ultimately yield an integer value. The next major difference is that when multiple case
statements yield the same value, they can all be combined into one line. SUGAR, PIZZA, and CHOCOLATE all yield 10, so we can simply write case SUGAR, PIZZA, CHOCOLATE -> 10;
.
Next, note that the break
statements are gone! The new arrow syntax replaces the need for both the :
after case
and the break
at the end of the case statement. The arrow signals that once the value is reached it is to be assigned to the tasteValue
variable and then stop. We no longer have to explicitly state the full assignment expression; just stating the value is enough. The JVM knows to set the integer value to tasteValue
.
We can still have a default
case at the end as a fallback option. Also notice that in this example the default
case does not return an integer, but instead throws an exception. In fact, there are three possibilities for what can come after the arrow:
- a value of the type the switch expression was declared with
- throw a new exception
- a code block that evaluates to a value of the correct type
One very important thing that you must keep in mind is that since switch expressions evaluate to a specific value of a specific type, you need to account for all possible cases.
If the data type is a primitive or an object, then you must provide a
default
case. The only exception is using anenum
because it is easier to account for every possibility.
Variations of switch expressions
You can also use colon case
statements in a switch expression. The only real difference in the code would be that a regular old switch
statement is assigned to a variable and there are no breaks. While this is a valid option, it is not preferable because it does not take advantage of the newer and more compact arrow syntax. It also makes it easier to lose track of whether you are looking at a switch
statement or a switch
expression because other than the variable assignment at the top and the absence of breaks, there are no visual indicators of that in the code.
Java 13 introduced the yield
keyword which can be used inside colon case
statements to identify the value the case
statement yields. It also replaces the break
statement and removes the need to explicitly mention the variable the value is assigned to. If you are going to use colons in your case
statements, using the new yield
keyword is the best option.
int tasteValue = switch (taste) {
case SUGAR:
case PIZZA:
case CHOCOLATE:
yield 10;
case MEATBALLS:
case STEAK:
yield 7;
case BROCCOLI:
yield 4;
case DIRT:
yield 1;
default:
throw new IllegalStateException("Invalid tastable object: " + taste);
};
The yield
keyword cannot be used inside a switch
statement. Likewise, break
cannot be used in a switch
expression. Therefore the use of yield
helps ensure that the reader of your code doesn’t forget which type of switch
they are reading.
There is also an in-between option that uses the new arrow syntax but puts a code block with a yield
in it to the right of the arrow. This might seem unnecessarily verbose compared to the first example of the arrow case syntax shown above, but there are situations in which this long way has some advantages. The yield
statement must be the last line in the code block, but you can call other functions in the lines before it. A simple example of this would be printing the value about to be yielded to the console right before yielding it, as you can see in the example below.
tasteValue = switch (taste) {
case SUGAR, PIZZA, CHOCOLATE -> {
System.out.println(10);
yield 10;
}
case MEATBALLS, STEAK -> {
System.out.println(7);
yield 7;
}
case BROCCOLI -> {
System.out.println(4);
yield 4;
}
case DIRT -> {
System.out.println(1);
yield 1;
}
default -> {
throw new IllegalStateException("Invalid tastable object: " + taste);
}
};
Conclusion
A switch expression can be used instead of a switch statement to make the code more concise and less error-prone. The entire switch expression is assigned to a variable because it yields a value. Unless you are using an enum
in your switch expression, you must include a default
case. You can yield a single value, throw an exception, or use a code block that ultimately evaluates to a single value.
In this topic we learned a few variations of switch expression syntax. The new arrow syntax allows us to put all of our cases that yield the same result on one line; the arrow replaces the colon and the break
. Java 13 introduced the yield
keyword, which can be used in switch expressions but not in switch statements. It can be used both with colon or arrow syntax, but it is typically used at the end of a code block to return a value, often after calling other functions earlier in the block. If yield
is used with colon syntax, it replaces the break
, same as the arrow does in arrow syntax. An easy way to differentiate between switch expressions and switch statements is that switch expressions cannot have break
in them, and switch statements cannot have yield
in them.