It seems that a simple extension property like the following does not work.
var Dog.age = 0;
What is the recommended way to implement this? I have tried the following, and it worked, but this will prevent any Dog object from cleaned up by the Garbage Collector, won't it?
class Dog
{
}
val dogAgeMap=HashMap<Dog, Int>();
var Dog.age:Int
get() = dogAgeMap[this]?: 0;
set(value){ dogAgeMap[this] = value}
class PetShop
{
fun work()
{
val d1 = Dog();
d1.age = 100;
val d2 = Dog();
d2.age = 200;
println("${d1.age}, ${d2.age}");
}
}
fun main(args:Array<String>)
{
PetShop().work();
}
Correct, this will prevent the Dog instances on which the age setter has been called to be GCed inside the scope of where the dogAgeMap is defined. If you defined the Dog.age extension property (and thus dogAgeMap) in a limited scope with a limited (short) lifespan, then you are okay.
However, if that is not the case, and you need the age info all across you application, then age should just be part of the original class definition and you don't ever run into this problem.
Solution in this case
class Dog(val age: Int)
If you need the age information only in one part of your application, then a better way would be to create the lookup (the HashMap) only for that part, or to simply use an enriched class with age (or a wrapper class with age) instead of the Dog class in that part of your application. And when you are done with work there, you clean up the map or the enriched class instances. In that way no instances will leak.
But if you really really want to do it with an extension property across the whole application, and thus you need to keep the reference to the dogAgeMap all the time, then you need to take care of leaking memory if you have a lot of instances that you go through and set their age.
If that is your case you can use a WeakHashMap<Dog, Int> instead. A WeakHashMap only keeps weak references and it won't prevent Dog instances to be GCed (once your strong references are no longer retained).
import java.util.WeakHashMap
val dogAgeMap = WeakHashMap<Dog, Int>()
var Dog.age: Int
get() = dogAgeMap[this] ?: 0
set(value) {
dogAgeMap[this] = value
}
Note however, that WeakHashMap is a Java class and not part of Kotlin core library, so if you use Kotlin for multiplatform, this won't work. In that case you would need a WeakHashMap implementation (library) on each platform.
An alternative way to do this if your data for dogs also contains an ID for each dog, would be to use the ID as the lookup key instead. That would be possible to port to all platforms. The implementation would then change to
// I am using a Long here, but it could be whatever type that
// is small enough to not cause memory concerns, since
// these keys would still exist in memory because a normal HashMap is used.
class Dog(val id: Long) {}
val dogAgeMap = HashMap<Long, Int>()
var Dog.age: Int
get() = dogAgeMap[id] ?: 0
set(value) {
dogAgeMap[id] = value
}
Related
I'm new to Kotlin and I'm trying to understand it, I've just written a simple example that shows how using data classes with maps is a bit tricky, because it seems to me that data classes have a strange behaviour. By default, they define hashCode() based on every property of the class. But they don't define a default equals() method.
This caused to me a lot of confusion because I created a HashMap with a Data Class as a key, but I didn't override hashCode() and equals(). My data class has a MutableList member. When I put an element in the map, I retrieved it using map.get(dataObject) as long as I didn't add an element to the MutableList. After that, even if the data object was still the same, and I found it using map.keys (map.keys.indexOf(dataObject) works), map.get(dataObject) failed, due to the hashCode().
I can fix it using a normal class or adding hashCode() and equals(), removing the MutableList from hashCode(), but I'm wondering if, due to the default behaviour, overriding hashCode() and equals() should be "mandatory" with data classes because otherwise using them with Maps can lead to errors.
Is there something else I can do to avoid this problem?
package cards
data class Player(val name: String, var cards: MutableList<Card>) {
constructor(name: String): this(name, mutableListOf())
//I don't need to define equals, so pointers are checked. But if I don't override hashCode, as it's based
//on every property, the hashCode is calculated considering the content of the MutableList!
// override fun hashCode(): Int {
// return name.hashCode()
// }
}
data class Card(val name: String, val suite: String)
class Game(val players: List<Player>) {
val cardMap: MutableMap<Player, MutableList<Card>> = mutableMapOf()
fun putIntoMapAndGiveCards() {
val newCards = cardMap.getOrDefault(players[0], mutableListOf())
newCards.add(Card(name = "Four", suite = "Clubs"))
cardMap[players[0]] = newCards
//This changes the default hashCode - I can use data classes in a list, but not in a map, because maps are
//based on it.
players[0].cards.add(Card(name = "Five", suite = "Clubs"))
}
fun getFromMap(): MutableList<Card>? {
val player = players[0]
assert(player != null, { "Player from list failure" })
val indexOfPlayer = cardMap.keys.indexOf(player)
assert(indexOfPlayer == 0, { "Player is in the map" })
//Without overriding hashCode, cards is null!
val cards = cardMap.get(players[0])
assert(cards != null, { "Cards from map failure" })
return cards
}
}
fun main() {
val player1 = Player(name = "John")
val game = Game(mutableListOf(player1))
game.putIntoMapAndGiveCards()
game.getFromMap()
?: throw Exception( """Map.get() failure because Player is a data class.
| A data class by default builds its hashCode with every property. As it contains a MutableList,
| the hashCode changes when I add elements to the list. This means that I can't find the element using get()
""".trimMargin())
println("Test finished!")
}
By default, they define hashCode() based on every property of the class. But they don't define a default equals() method
This is not correct. Data classes generate both equals() and hashCode() consistently based on the properties declared in the data class's primary constructor (same goes for toString() btw).
Here is the decompiled code for equals and hashCode of your Player class:
public int hashCode() {
String var10000 = this.name;
int var1 = (var10000 != null ? var10000.hashCode() : 0) * 31;
List var10001 = this.cards;
return var1 + (var10001 != null ? var10001.hashCode() : 0);
}
public boolean equals(#Nullable Object var1) {
if (this != var1) {
if (var1 instanceof Player) {
Player var2 = (Player)var1;
if (Intrinsics.areEqual(this.name, var2.name) && Intrinsics.areEqual(this.cards, var2.cards)) {
return true;
}
}
return false;
} else {
return true;
}
}
Your problem is that you declare your cards mutable list in the primary constructor so it's part of the generated equals and hashCode.
The solution is to move this cards property to the body of your class instead (since it's not part of the player's "core data", but rather part of the state):
data class Player(val name: String) {
val cards: MutableList<Card> = mutableListOf()
}
This way, the generated equals/hashCode pair will only be based on the name property.
Another option obviously is to override both equals and hashCode manually to take only the name into account, but that's tedious and not very idiomatic.
I'm wondering if, due to the default behaviour, overriding hashCode() and equals() should be "mandatory" with data classes because otherwise using them with Maps can lead to errors.
I think you have misdiagnosed the default behaviour. So I'd say on the contrary overriding equals/hashCode is actually not very idiomatic for data classes, and should in general be avoided.
Using data classes is usually safe in maps, as long as the data in the primary constructor is not mutable.
Side notes
you really should not mix var with mutable collections. It creates 2 ways of changing the collection, which is pretty unexpected and error-prone. You should instead either use a val MutableList or a var List, so you can only change the list via mutation, or only change it via assignment, but not both.
if you want to insert the new value into the map, you shouldn't use getOrDefault + assign the value to the key. Instead, use getOrPut directly, so the default value will be inserted without extra work.
why are you both using a cards property on the Player and a Map<Player, List<Card>>? Looks like you have 2 states that can change independently now because those card lists are independent.
I am developing a simple Android app, that will display an icon of a vehicle and the user can click on the icon to display the vehicle information. I want to load the data dynamically when I build the app i.e. the data will come from an external source including the picture for the icon.
I am new to Kotlin and not sure what to search for to understand a suitable solution. What is the correct way to define the data, is it best to create an class as below then create an array of the class (not sure if this is possible)
public class VehicleSpec()
{
var OEM: String? = null
var ModelName: String? = null
var EngineSize: String? = null
}
Or would be better to create a multiple dimension array and then link the data to the cells?
var VehicleSpec = arrayOf(20,20)
VehicleSpec[0][0] = Null //OEM
VehicleSpec[0][1] = Null //ModelName
VehicleSpec[0][2] = Null //EngineSize
What is the best way to set up the data storage, is there any good references to understand how this should be setup?
What is the correct way to define the data, is it best to create an class as below then create an array of the class
Using an array for the properties of an object is not making the full use of the type safety you have in Kotlin (and even Java for that matter).
If what you want to express is multiple properties of an object, then you should use a class to define those properties. This is especially true if the properties have different types.
There is no performance difference between an array and a class, because you'll get a reference to the heap in both cases. You could save on performance only if you convert your multi-dimensional array approach to a single-dimension array with smart indexing. Most of the time, you should not consider this option unless you are handling a lot of data and if you know that performance is an issue at this specific level.
(not sure if this is possible)
Defining lists/arrays of classes is definitely possible.
Usually, for classes that are only used as data containers, you should prefer data classes, because they give you useful methods for free, and these methods totally make sense for simple "data bags" like in your case (equals, hashcode, component access, etc.).
data class Vehicle(
val OEM: String,
val ModelName: String,
val EngineSize: String
)
Also, I suggest using val instead of var as much as possible. Immutability is more idiomatic in Kotlin.
Last but not least, prefer non-null values to null values if you know a value must always be present. If there are valid cases where the value is absent, you should use null instead of a placeholder value like empty string or -1.
First at all, using the "class aprocah" makes it easy for you to understand and give you the full benefits of the language itself... so dont dry to save data in an array .. let the compiler handle those stuff.
Secondly i suggest you have maybe two types (and use data classes ;-) )
data class VehicleListEntry(
val id: Long,
val name: String
)
and
data class VehicleSpec(
val id: Long,
val oem: String = "",
val modelName: String = "",
val engineSize: String = ""
)
from my perspective try to avoid null values whenever possible.
So if you have strings - which you are display only - use empty strings instead of null.
and now have a Model to store your data
class VehicleModel() {
private val specs: MutableMap<Long, VehicleSpec> = mutableMapOf()
private var entries: List<VehicleListEntry> = listOf()
fun getSpec(id: Long) = specs[id]
fun addSpec(spec: VehicleSpec) = specs[spec.id] = spec
fun getEntries(): List<VehicleListEntry> = entries
fun setEntries(data: List<VehicleListEntry>) {
entries = data.toMutableList()
}
}
You could also use a data class for your model which looks like
data class VehicleModel(
val specs: MutableMap<Long, VehicleSpec> = mutableMapOf(),
var entries: List<VehicleListEntry> = listOf()
)
And last but not least a controller for getting stuff together
class VehicleController() {
private val model = VehicleModel()
init{
// TODO get the entries list together
}
fun getEntries() = model.entries
fun getSpec(id: Long) : VehicleSpec? {
// TODO load the data from external source (or check the model first)
// TODO store the data into the model
// TODO return result
}
}
I have an immutable object:
class Foo(
val name: String,
val things: List<Thing>
)
A third party lib creates the Foo object with some 'null' Thing objects.
I am creating a new object:
val foo = thirdPartyGetFoo()
val filteredFoo = Foo(foo.name, foo.things.filterNotNull())
That works, however AndroidStudio greys out the filterNotNull function call and presents a warning:
Useless call on collection type: The inspection reports filter-like
calls on already filtered collections.
Is this the right way to filter that list? Should I ignore the warning or is there a better way?
You do not specify what library creates the object with nulls. Some deserialization libraries can use static factory methods which you could configure, and then have the factory method strip the null. For example, if this were Jackson you would simply:
class Foo(val name: String, val things: List<Thing>) {
companion object {
#JsonCreator
#JvmName("createFromNullable")
fun create(name: String, things: List<Thing?>) = Foo(name, things.filterNotNull())
fun create(name: String, things: List<Thing>) = Foo(name, things)
}
}
Then...
val goodFoo = jacksonObjectMapper().readValue<Foo>(someJsonWithNulls)
Maybe your library has options that are similar?
If not, and you don't have 100 of these things with this problem, I would probably create a temporary class to hold the results and convert that to the final class:
open class FooNullable(val name: String, open val things: List<Thing?>) {
open fun withoutNulls(): Foo = Foo(name, things.filterNotNull())
}
class Foo(name: String, override val things: List<Thing>) : FooNullable(name, things) {
override fun withoutNulls(): Foo = this
}
Then you can deserialize into FooNullable and just call withoutNulls() to get the other flavor that is clean. And if you accidentally call it on one without nulls already, it just does nothing.
val goodFoo = Foo("", emptyList<Thing>())
val alsoGoodFoo = goodFoo.withoutNulls() // NOOP does nothing
val badFoo = thirdPartyGetFoo()
val betterFoo = badFoo.withoutNulls() // clean up the instance
val safeFoo = thirdPartyGetFoo().withoutNulls() // all at once!
Not the cleanest, but does work. The downsides is this second step, although it looks like you were already planning on doing that anyway. But this model is safer than what you proposed since you KNOW which type of object you have and therefore you continue to be typesafe and have the compiler helping you avoid a mistake.
You don't have to use inheritance as in the above example, I was just trying to unify the API in case there was a reason to have either version in hand and know which is which, and also act upon them in a similar way.
I want to set up two values that hold immutable references to each other. Example:
data class Person(val other: Person)
val jack = Person(jill), jill = Person(jack) // doesn't compile
Note: lateinit doesn't seem to work with data class primary constructors.
Any ideas?
You could get away with something like this:
class Person() {
private var _other: Person? = null
private constructor(_other: Person? = null) : this() {
this._other = _other
}
val other: Person
get() {
if (_other == null) {
_other = Person(this)
}
return _other ?: throw AssertionError("Set to null by another thread")
}
}
And then you would be able to do:
val jack = Person()
val jill = jack.other
Using a data class here does not work for multiple reasons:
First because a data class can't have an empty constructor.
Even if that wasn't a problem, the generated methods would end up having a cyclic dependency and will fail in runtime with java.lang.StackOverflowError. So you'd have to overwrite toString, equals, etc. which kind of defeats the purpose of using data class in the first place.
Here is the trick (note, this is really a trick, you need a good reason to use it in real code).
Unfortunately it won't work with data classes, as they seem to be secured against this kind of hacks.
But if you have java-stile classes, you may use two things to your advantage:
You can initialize vals in the constructor (same as with final in java)
You have access to this inside the constructor (and you may leak it outside if you really want)
Which means that you can create another Person inside the constructor of the first person and finalize the creation of both classes before the constructor finishes.
Once again: exposing this as I did below is a bad idea. When otherFactory is called, it's parameter is only half-initialized. This may lead to nasty bugs, especially if you try to publish such reference in multithreaded environment.
A bit safer approach is to create both Persons inside the constructor of the first Person (you'll need to supply the fields of both entities as arguments). It's safer because you're in control of the code that uses half-initialized this reference.
class Person {
val name: String
val other: Person
constructor(name: String, other: Person) {
this.name = name
this.other = other
}
// !! not very safe !!
constructor(name: String, otherFactory: (Person) -> Person) {
this.other = otherFactory(this)
this.name = name
}
// a bit safer
constructor(name: String, otherName: String) {
this.other = Person(otherName, this)
this.name = name
}
}
val person1 = Person("first") {
Person("second", it)
}
val person2 = person1.other
print(person1.name) // first
print(person2.name) // second
val person3 = Person("third", "fourth")
val person4 = person3.other
print(person3.name)
print(person4.name)
Thanks for your suggestions everybody. I came up with an alternative and would like to hear your insights:
open class Person {
open val other: Person by lazy { Person2(this) }
class Person2(override val other: Person): Person()
}
val jack = Person()
val jill = jack.other
Here we have one person lazily instantiating the other on demand, using an internal subclass that implements other differently (i.e. it is just given it directly in its constructor).
Thoughts most welcome.
It's easy to write extension methods in Kotlin:
class A { }
class B {
fun A.newFunction() { ... }
}
But is there some way to create extension variable? Like:
class B {
var A.someCounter: Int = 0
}
You can create an extension property with overridden getter and setter:
var A.someProperty: Int
get() = /* return something */
set(value) { /* do something */ }
But you cannot create an extension property with a backing field because you cannot add a field to an existing class.
No - the documentation explains this:
Extensions do not actually modify classes they extend. By defining an extension, you do not insert new members into a class, but merely make new functions callable with the dot-notation on instances of this class.
and
Note that, since extensions do not actually insert members into classes, there’s no efficient way for an extension property to have a backing field. This is why initializers are not allowed for extension properties. Their behavior can only be defined by explicitly providing getters/setters.
Thinking about extension functions/properties as just syntactic sugar for calling a static function and passing in a value hopefully makes this clear.
However, if you really, really want to do something like this...
As stated above regarding efficiency, an additional backing field added directly to the class is the best way to store data non-derivable from existing non-private members from the class. However, if you don't control the implementation of the class and are dead-set on creating a new property that can store new data, it can be done in a way that is not abysmally inefficient by using separate external tables. Use a separate map that keys on object instances of this class with values that map directly to the value you want to add then define an extension getter and/or setter for this property which uses your external table to store the data associated with each instance.
val externalMap = mutableMapOf<ExistingClass, Int>()
var ExistingClass.newExtensionProperty : Int
get() = externalMap[this] ?: 0
set(value:Int) { externalMap[this] = value }
The additional map lookups will cost you - and you need to consider memory leaks, or using appropriately GC-aware types, but it does work.
There's no way to add extension properties with backing fields to classes, because extensions do not actually modify a class.
You can only define an extension property with custom getter (and setter for var) or a delegated property.
However, if you need to define an extension property which would behave as if it had a backing field, delegated properties come in handy.
The idea is to create a property delegate that would store the object-to-value mapping:
using the identity, not equals()/hashCode(), to actually store values for each object, like IdentityHashMap does;
not preventing the key objects from being garbage collected (using weak references), like WeakHashMap does.
Unfortunately, there is no WeakIdentityHashMap in JDK, so you have to implement your own (or take a complete implementation).
Then, based on this mapping you can create a delegate class satisfying the property delegates requirements. Here's an example non-thread-safe implementation:
class FieldProperty<R, T : Any>(
val initializer: (R) -> T = { throw IllegalStateException("Not initialized.") }
) {
private val map = WeakIdentityHashMap<R, T>()
operator fun getValue(thisRef: R, property: KProperty<*>): T =
map[thisRef] ?: setValue(thisRef, property, initializer(thisRef))
operator fun setValue(thisRef: R, property: KProperty<*>, value: T): T {
map[thisRef] = value
return value
}
}
Usage example:
var Int.tag: String by FieldProperty { "$it" }
fun main(args: Array<String>) {
val x = 0
println(x.tag) // 0
val z = 1
println(z.tag) // 1
x.tag = "my tag"
z.tag = x.tag
println(z.tag) // my tag
}
When defined inside a class, the mapping can be stored independently for instances of the class or in a shared delegate object:
private val bATag = FieldProperty<Int, String> { "$it" }
class B() {
var A.someCounter: Int by FieldProperty { 0 } // independent for each instance of B
var A.tag: String by bATag // shared between the instances, but usable only inside B
}
Also, please note that identity is not guaranteed for Java's primitive types due to boxing.
And I suspect the performance of this solution to be significantly worse than that of regular fields, most probably close to normal Map, but that needs further testing.
For nullable properties support and thread-safe implementation please refer to here.
You can't add a field, but you can add a property, that delegates to other properties/methods of the object to implement its accessor(s). For example suppose you want to add a secondsSinceEpoch property to the java.util.Date class, you can write
var Date.secondsSinceEpoch: Long
get() = this.time / 1000
set(value) {
this.time = value * 1000
}
If you are extending View you can do it quite easily like this...
This is example how I create some my custom class Event property in EditText class extension:
Define id for key :
<?xml version="1.0" encoding="utf-8"?>
<resources>
<item name="EditTextEventOnClearTagKey" type="id" />
</resources>
Define one reusable extension like this:
fun <T : Any> View.tagProperty(#IdRes key: Int, onCreate: () -> T): T {
#Suppress("UNCHECKED_CAST")
var value = getTag(key) as? T
if (value.isNull) {
value = onCreate()
setTag(key, value)
}
return value!!
}
Use it in wherever View extension you need:
val EditText.eventClear get() = tagProperty(R.id.EditTextEventOnClearTagKey) { event<Unit>() }