As kotlin reference Classes and Inheritance say,
If the class has a primary constructor, each secondary constructor needs to delegate to the primary
This is because init
blocks and property initializers always need to run properly to construct an instance of a class, and they might rely on properties passed to the primary constructor to do their initialization - this is the convenience that the primary constructor gives you (as well as being able to have properties right in the header of the class).
As an example, take this class:
class Rectangle(val width: Int, val height: Int) {
constructor(size: Int) : this(size, size)
val area = width * height
init {
println("New rectangle, $width x $height")
}
}
Both the area
property and the init
block make use of primary constructor parameters - if the secondary constructor didn't call through to the primary one, initialization couldn't be performed.
The width
and height
properties are also implicitly initialized when the primary constructor is called - again, if the secondary constructor didn't call the primary, these would be left uninitialized.
Of course, you can have multiple secondary constructors in a class if there's no primary constructor (this is common for Android Views for example) - you'll just have a harder time performing your initialization logic if there's any.