I want to add a delegated property to my Composition class, as follows
#Entity
data class Composition(
val author: String="",
val background: String?,
val description: String="",
val downloadURL: String="",
val duration: String="",
val headPortrait: String?,
#PrimaryKey val id: String,
val isLike: Boolean,
val likeAmount: String="",
val playingAmount: Int=0,
val replyAmount: String?,
val showStyle: String?,
val title: String?,
val userId: String?,
val commentAmount: String?,
val cover: String=""
){
val showDuration by lazy{
val minutes = duration.toInt() /60
val seconds =duration.toInt()%60
"$minutes:$seconds"
}
}
But there gona be a compile error because delegated property can not be saved in database. So i want to
add an
Ignore annotation to this field. What a pity that Androidstuio will throws a complain "This annotation is not applicable to target 'member property with delegate" . Who has ideas for this problem?
As user IR42 points out in his comment, you can use #delegate:Ignore.
Related
I am trying to implement a room database for a league data but when I'm trying to compile, it's doesnt work. The error seem to be throw in the Database file. I hope you can help me
In the Dao, i select all the countryProperty and insert them in the database,
#Dao
interface CountryDataBaseDao {
#Query("SELECT * FROM DatabaseCountryProperty")
fun getData() : LiveData<List<DatabaseCountryProperty>>
#Insert(onConflict = OnConflictStrategy.REPLACE)
fun insertData(data : List<DatabaseCountryProperty>)
}
In the entity, countryProperty entity get the extra entity,
#Entity
data class DatabaseCountryProperty constructor(
#PrimaryKey
val id: String,
val name: String,
val imagePath: String?,
val extra: DatabaseExtraCountry?)
data class DatabaseExtraCountry constructor(
val continent : String?,
val subRegion : String?,
val worldRegion : String?,
val fifa : String?,
val iso : String?,
val iso2 : String?,
val longitude : Double?,
val latitude : Double?,
val flag : String?) {
}
fun List<DatabaseCountryProperty>.asDomainModel() : List<DevByteCountryProperty> {
return map {
DevByteCountryProperty(
id = it.id,
name = it.name,
imagePath = it.imagePath,
extra = DevByteExtraCountry(
continent = it.extra?.continent,
subRegion = it.extra?.subRegion,
worldRegion = it.extra?.worldRegion,
fifa = it.extra?.fifa,
iso = it.extra?.iso,
iso2 = it.extra?.iso2,
longitude = it.extra?.longitude,
latitude = it.extra?.latitude,
flag = it.extra?.flag
)
)
}
}
In the Database,
#Database(entities = [DatabaseCountryProperty::class], version = 1, exportSchema = false)
abstract class DataDataBase : RoomDatabase() {
abstract val countryDataBaseDao : CountryDataBaseDao
}
private lateinit var INSTANCE: DataDataBase
fun getDatabase(context: Context): DataDataBase {
synchronized(DataDataBase::class.java) {
if (!::INSTANCE.isInitialized) {
INSTANCE = Room.databaseBuilder(context.applicationContext,
DataDataBase::class.java,
"data").build()
}
}
return INSTANCE
}
The errors was that I didn't had the annotation embedded to declare the fact that the extra is a sub_class. So the code is,
#Entity
data class DatabaseCountryProperty constructor(
#PrimaryKey
val id: String,
val name: String,
val imagePath: String?,
#Embedded
val extra: DatabaseExtraCountry?)
data class DatabaseExtraCountry constructor(
val continent : String?,
val subRegion : String?,
val worldRegion : String?,
val fifa : String?,
val iso : String?,
val iso2 : String?,
val longitude : Double?,
val latitude : Double?,
val flag : String?) {
}
Add these in your build gradle.
id 'kotlin-kapt' in your plugin
kapt "androidx.room:room-compiler:2.3.0" in your dependencies
Just Do a Try
I'm try to convert a sqs message to a self defined SQS event, the message in convertMessageFromSqsForEvent is valid and it have value, but the sqsEventProperties is null, what is the problem?
fun convertMessageFromSqsForEvent(message: Message): SQSEventProperties? {
var sqsEventProperties: SQSEventProperties? = null
sqsEventProperties = mapper?.readValue(message.body, SQSEventProperties::class.java)
return sqsEventProperties
}
data class SQSEventProperties(
#JsonProperty("attributes") var attributes: String?,
#JsonProperty("awsRegion") var awsRegion: String?,
#JsonProperty("body") var body: String?,
#JsonProperty("eventSource") var eventSource: String?,
#JsonProperty("eventSourceARN") var eventSourceARN: String?,
#JsonProperty("md5OfBody") var md5OfBody: String?,
#JsonProperty("md5OfMessageAttributes") var md5OfMessageAttributes: String?,
#JsonProperty("messageId") var messageId: String?,
#JsonProperty("receiptHandle") var receiptHandle: String?,
)
message
{MessageId: 94995ca7-ee33-4109-aabf-23d375878140,
ReceiptHandle: AQEBf5kvP0RwvFtKPmZglRhxXAS7FKGxbUCazYXD0+HFq6hpQGri3KU8TZirGC9NEu86DZretaBi5oeFbzzkWnMoxwzZ68/m7FZLt8fxX0gCoL2CxJMy0JXClRTZI+O06Hwn+CZMb5LOAKkGffJUS48dfe5GERKAZGsnLK4vA2pODrxfv1vvC6CXTzejoA9Dw+kuo51F5S86iRzXjItjzRJDxcJp4xeXdjrWps4Wfx233G+x4KiP9t9yZ73L9ucDLzdguDwTee07KG5SdRgHDQWyXcc0IJNtAe/NDmkIhWhigHoZSQmVCVNQFBtcYuB6X2khB+oQZsLz0Vh1NLs3zQVuusBbMg6tpk94N28FUKCIugtwDxBRjIdJqfbUXEeL1PxXCBcb+Pn3vLH9NHKHEclxbg==,
MD5OfBody: de5da762733bf71530cdb2aba2892a65,
Body: Please send request for updated data,
Attributes: {},
MessageAttributes: {}}
The name of the field of the data class must be the same as the key of the Json data. You can modify your data class and try again.
data class SQSEventProperties(
#JsonProperty("Attributes") var attributes: String?,
#JsonProperty("AwsRegion") var awsRegion: String?,
#JsonProperty("Body") var body: String?,
#JsonProperty("EventSource") var eventSource: String?,
#JsonProperty("EventSourceARN") var eventSourceARN: String?,
#JsonProperty("Md5OfBody") var md5OfBody: String?,
#JsonProperty("Md5OfMessageAttributes") var md5OfMessageAttributes: String?,
#JsonProperty("MessageId") var messageId: String?,
#JsonProperty("ReceiptHandle") var receiptHandle: String?,
)
So I have a sample app that is currently using Retrofit to fetch data and display it in a recyclerview with its custom adapter. I want to pass the data to the a more details page when I click on a character name on my recyclerView. I found some tutorials and decided to use the Parcelize annotation like this:
#Parcelize data class Character (
val charID: Long,
val name: String,
val birthday: String,
val occupation: List<String>,
val img: String,
val status: String,
val nickname: String,
val appearance: List<Long>,
val portrayed: String,
val category: String,
val betterCallSaulAppearance: List<Long> ) : Parcelable
My adapter looks like this:
override fun onBindViewHolder(holder: ViewHolder, position: Int) {
val character = characters[position]
holder.characterNameText.text = character.name
holder.characterNameText.setOnClickListener {
holder.passData(character)
}
}
override fun getItemCount() = characters.size
class ViewHolder(itemView : View) : RecyclerView.ViewHolder(itemView) {
val characterNameText = itemView.findViewById<TextView>(R.id.character_name)
fun passData(character : Character) {
val intent = Intent(itemView.context, CharacterDetails::class.java)
intent.putExtra(CHARACTER, character)
itemView.context.startActivity(intent)
}
}
And in the CharacterDetails Activity it looks like this:
companion object {
const val CHARACTER = "character"
}
override fun onCreate(savedInstanceState: Bundle?) {
super.onCreate(savedInstanceState)
setContentView(R.layout.activity_character_details)
val bundle = intent.extras
val character = bundle!!.getParcelable<Character>(CHARACTER)
val characterName = findViewById<TextView>(R.id.character_details_name)
Yet I get a
java.lang.NullPointerException: Attempt to invoke interface method 'int java.util.List.size()' on a null object reference
at com.example.myapplication.models.Character.writeToParcel(Unknown Source:85)
I'm still new to this so I really need your help.Thanks!
Obviously, the error caused by the list reference of the data class is null. You can modify the code like this.
#Parcelize data class Character (
val charID: Long,
val name: String,
val birthday: String,
val occupation: List<String>?,
val img: String,
val status: String,
val nickname: String,
val appearance: List<Long>?,
val portrayed: String,
val category: String,
val betterCallSaulAppearance: List<Long>? ) : Parcelable
I have the following two data classes
data class CreateMedicDto(
val firstName: String,
val lastName: String,
val pin: String,
val address: String?,
val birthDate: LocalDate,
val specialty: Specialty,
)
data class UpdateMedicDto(
val firstName: String,
val lastName: String,
val address: String?,
val birthDate: LocalDate,
val specialty: Specialty,
)
The only difference is that the second one is missing pin field. The reason for that is that in case of update I do not want to allow the possibility to change the pin by using making use of the language and framework features.
Currently, in this form, I will need to validate both of them:
fun validateMedic(
input: CreateMedicDto,
): MedicValidationResult? {
with(input) {
if (checkLengthBetween1And50(input.firstName)) return MedicValidationResult.InvalidFirstNameLength
if (checkLengthBetween1And50(input.lastName)) return MedicValidationResult.InvalidLastNameLength
address?.let {
if (checkLengthGreaterThan500(it)) return MedicValidationResult.InvalidAddressLength
}
if (checkDateValidity(birthDate)) return MedicValidationResult.InvalidBirthDate
}
return null
}
fun validateMedic(
input: UpdateMedicDto,
): MedicValidationResult? {
with(input) {
if (checkLengthBetween1And50(input.firstName)) return MedicValidationResult.InvalidFirstNameLength
if (checkLengthBetween1And50(input.lastName)) return MedicValidationResult.InvalidLastNameLength
address?.let {
if (checkLengthGreaterThan500(it)) return MedicValidationResult.InvalidAddressLength
}
if (checkDateValidity(birthDate)) return MedicValidationResult.InvalidBirthDate
}
return null
}
The code is almost identical, only the input parameter type is different.
The question is how can I reduce it to a single function?
I have some solutions but there are also some reasons for that I would not like to use them:
Inherit CreateMedicDto from UpdateMedicDto. This may easily solve the problem but it doesn't really make much sense to say that CreateMedicDto is an UpdateMedicDto
Make that pin field nullable. With this approach I abandon the language null-safety feature and I am going to rely on a parameter which says "create" or "update"
Both extending a common class. The third class is actually useless and can't find a proper name for it
Any better approaches to still benefit from the type safety?
Your option 3 is called an abstract class. abstract means it cannot be instantiated, but also that it can have abstract members (properties and functions with no implementation or initial value).
An option 4 would be to create a common interface for both of them.
But in both cases, there will still be duplication of all the property names because you have to override them. But it fixes the duplication at your validation functions.
Option 3 example:
abstract class MedicDto {
abstract val firstName: String
abstract val lastName: String
abstract val pin: String
abstract val address: String?
abstract val birthDate: LocalDate
abstract val specialty: Specialty
}
data class CreateMedicDto(
override val firstName: String,
override val lastName: String,
val pin: String,
override val address: String?,
override val birthDate: LocalDate,
override val specialty: Specialty,
): MedicDto()
data class UpdateMedicDto(
override val firstName: String,
override val lastName: String,
override val address: String?,
override val birthDate: LocalDate,
override val specialty: Specialty,
): MedicDto()
Option 4 example:
interface MedicDto {
val firstName: String
val lastName: String
val pin: String
val address: String?
val birthDate: LocalDate
val specialty: Specialty
}
data class CreateMedicDto(
override val firstName: String,
override val lastName: String,
val pin: String,
override val address: String?,
override val birthDate: LocalDate,
override val specialty: Specialty,
): MedicDto
data class UpdateMedicDto(
override val firstName: String,
override val lastName: String,
override val address: String?,
override val birthDate: LocalDate,
override val specialty: Specialty,
): MedicDto
I am using Kotlin 1.30. I have the following entity classes that will be populated from the API. And some of the properties could be null from the API so I have declared them using the safe null type.
However, I will map this entity class to my domain model class. And just wondering what is the best way to handle the null types?
I have 2 ideas on how to do this when I map the classes.
declare all the equivalent properties as safe null types
use the elivs operator to return either a empty string or a emptylist
In the following snippet I am using the elvis operator. Just wondering what is the best practice for this?
class LoginResponseDomainMapperImp : LoginResponseDomainMapper {
override fun map(entity: LoginResponseEntity): LoginResponse {
return LoginResponse(
entity.token ?: "",
mapUsers(entity.user),
mapEnterprises(entity.enterprises ?: emptyList()),
mapVendors(entity.vendors ?: emptyList()))
}
private fun mapUsers(userEntity: UserEntity?): User {
return User(
userEntity?.id,
userEntity?.email,
userEntity?.firstName,
userEntity?.lastName,
userEntity?.phone,
userEntity?.address,
userEntity?.dob,
userEntity?.customer,
userEntity?.enterpriseIds ?: emptyList(),
userEntity?.vendorIds ?: emptyList())
}
private fun mapEnterprises(enterprisesEntity: List<EnterprisesEntity>): List<Enterprises> {
val enterpriseList = mutableListOf<Enterprises>()
enterprisesEntity.forEach {
val enterprise = Enterprises(
it.id,
it.enterpriseName,
it.typeLabel,
it.country,
it.imageId,
it.managers,
it.members,
it.stripe,
it.locations)
enterpriseList.add(enterprise)
}
return enterpriseList.toList()
}
private fun mapVendors(vendorsEntity: List<VendorsEntity>): List<Vendors> {
val vendorList = mutableListOf<Vendors>()
vendorsEntity.forEach {
val vendor = Vendors(
it.id,
it.vendorName,
it.typeLabel,
it.userRole,
it.latitude,
it.longitude,
it.partner,
it.country,
it.imageId,
it.stripe)
vendorList.add(vendor)
}
return vendorList.toList()
}
}
Entity class that will populate from the API, so any of these could be null
data class LoginResponseEntity(
#SerializedName("token") val token: String?,
#SerializedName("user") val user: UserEntity?,
#SerializedName("enterprises") val enterprises: List<EnterprisesEntity>?,
#SerializedName("vendors") val vendors: List<VendorsEntity>?)
data class UserEntity(
#SerializedName("id") val id: String?,
#SerializedName("email") val email: String?,
#SerializedName("first_name") val firstName: String?,
#SerializedName("last_name") val lastName: String?,
#SerializedName("phone") val phone: String?,
#SerializedName("address") val address: String?,
#SerializedName("dob") val dob: String?,
#SerializedName("customer") val customer: String?,
#SerializedName("enterprise_ids") val enterpriseIds: List<String>?,
#SerializedName("vendor_ids") val vendorIds: List<String>?)
data class EnterprisesEntity(
#SerializedName("id") val id: String?,
#SerializedName("enterprise_name") val enterpriseName: String?,
#SerializedName("type_label") val typeLabel: String?,
#SerializedName("referral_code") val referralCode: String?,
#SerializedName("country") val country: String?,
#SerializedName("image_id") val imageId: String?,
#SerializedName("managers") val managers: List<String>?,
#SerializedName("members") val members: List<String>?,
#SerializedName("stripe") val stripe: Boolean,
#SerializedName("locations") val locations: List<String>?)
data class VendorsEntity(
#SerializedName("id") val id: String?,
#SerializedName("vendor_name") val vendorName: String?,
#SerializedName("type_label") val typeLabel: String?,
#SerializedName("user_role") val userRole: String?,
#SerializedName("latitude") val latitude: Float,
#SerializedName("longitude") val longitude: Float,
#SerializedName("partner") val partner: Boolean,
#SerializedName("country") val country: String?,
#SerializedName("image_id") val imageId: String?,
#SerializedName("stripe") val stripe: Boolean)
Data model class in the domain, Its it better to declare them all safe null types?
data class LoginResponse(
val token: String,
val user: User?,
val enterprises: List<Enterprises>,
val vendors: List<Vendors>)
data class User(
val id: String?,
val email: String?,
val firstName: String?,
val lastName: String?,
val phone: String?,
val address: String?,
val dob: String?,
val customer: String?,
val enterpriseIds: List<String>,
val vendorIds: List<String>)
data class Enterprises(
val id: String,
val enterpriseName: String,
val typeLabel: String,
val country: String,
val imageId: String,
val managers: List<String>,
val members: List<String>,
val stripe: Boolean,
val locations: List<String>)
data class Vendors(
val id: String,
val vendorName: String,
val typeLabel: String?,
val userRole: String,
val latitude: Float,
val longitude: Float,
val partner: Boolean,
val country: String?,
val imageId: String,
val stripe: Boolean)
First of all there are no safe null types in Kotlin. A variable is either nullable or not.
If the API delivers null values, either by absence or by setting them explicitely null, your data classes should reflect that in the way that those variables are nullable (question mark ? right after the data type).
The moment you map those data classes to your entities (domain model) you should handle the null case properly. Since it is a valid case, that the API serves null values, which you expect, you should handle that case by assigning a default value.
Using the elvis operator is just a way to deal with nullable types elegantely, but if you use it, remains your choice.
If you declare all the properties as nullable that's not much better than Java, even when you access them with the null safety operator. If your JSON properties have null value, that means that in your business logic they don't always have a value, and is your responsibility to handle that, by fallback to a default value (i.e. emptyList() or ""), or maybe something more complicated like redirecting to a sign in screen.