graphql-provider/User implementation fields
In actual projects, in addition to the fields that can be automatically mapped as discussed earlier, entity objects should also have user-defined fields whose functions are determined by the business and should also be implemented by users.
In this article we demonstrate two user-implemented fields
- Author.fullName
This is a simple user implementation field that requires neither access the database nor batch load using the DataLoader
- BookStore.avgPrice
This is a complex user implementation field that requires not only access to the database, but also batch loading using DataLoader
-
Add new kotlin property in entity interface
interface Author { ...other properties... val fullName: String }
-
Change AuthorMapper
@Component class AuthorMapper: EntityMapper<Author, UUID> { override fun EntityTypeDSL<Author, UUID>.config() { ...other configuration... userImplementation(Author::fullName) // α } fun fullName(separator: String = " ") = // β runtime.implement(Author::fullName) { // γ "${it.firstName}${separator}${it.lastName}" // δ } }
-
α
Map Author.fullName as user implementation field
-
β
Provides a public function.
-
In theory, the function name is arbitrary, but for readability it is recommended to keep the same as the name of Author::fullName at γ
-
The arguments of this function will be converted to the arguments of the Author.fullName field in the GraphQL Schema
-
This function has no return type
-
-
γ
-
runtime is a protected property provided by the superclass org.babyfish.graphql.provider.EntityMapper
-
Auhor::fullName represents the field for which you want to implement by your code
-
-
δ
it is an implicit variable provided by the lambda expression of the runtime.implement function, representing the current Author object
-
Start the app, access http://localhost:8080/graphiql and execute
query {
books {
name
store {
name
}
authors {
fullName
}
}
}
The response is
{
"data": {
"books": [
{
"name": "Effective TypeScript",
"store": {
"name": "O'REILLY"
},
"authors": [
{
"fullName": "Dan Vanderkam"
}
]
},
{
"name": "GraphQL in Action",
"store": {
"name": "MANNING"
},
"authors": [
{
"fullName": "Samer Buna"
}
]
},
{
"name": "Learning GraphQL",
"store": {
"name": "O'REILLY"
},
"authors": [
{
"fullName": "Eve Procello"
},
{
"fullName": "Alex Banks"
}
]
},
{
"name": "Programming TypeScript",
"store": {
"name": "O'REILLY"
},
"authors": [
{
"fullName": "Boris Cherny"
}
]
}
]
}
}
-
Add new kotlin property in entity interface
interface BookStore { ...other properties... val avgPrice: BigDecimal }
-
Add your repository
Create new package com.example.demo.dal, add BookRepository under it
package com.example.demo.dal import com.example.demo.model.Book import com.example.demo.model.price import com.example.demo.model.store import org.babyfish.graphql.provider.runtime.R2dbcClient import org.babyfish.kimmer.sql.ast.avg import org.babyfish.kimmer.sql.ast.valueIn import org.springframework.stereotype.Repository import java.math.BigDecimal import java.util.* @Repository // α class BookRepository( private val r2dbcClient: R2dbcClient // β ) { suspend fun findAvgPriceGroupByStoreIds( storeIds: Collection<UUID> // γ ): Map<UUID, BigDecimal> = // δ r2dbcClient.query(Book::class) { select { where(table.store.id valueIn storeIds) groupBy(table.store.id) table.store.id then table.price.avg().asNonNull() } }.associateBy({it.first}) { // ε it.second } }
-
α
The current object needs to be managed by spring
-
β
Inject org.babyfish.graphql.provider.runtime.R2dbcClient
org.babyfish.graphql.provider.runtime.R2dbcClient is a wrapper for SqlClient of kimmer-sql. It combines kimmer-sql with spring data r2dbc, allowing kimmer-sql to enjoy the connection management and transaction management of spring data r2dbc.
-
γ
This query function accepts many BookStore ids
-
δ
For each bookstore id specified by the caller, return the average price of all books in the bookstore.
-
ε
The query result of strongly typed SQL contains two columns
- First column: group key, bookstore id
- The second column: the average price of all books in the current group
The associateBy function here uses the first column of the query result as the key and the second column of the query result as the value, and returns a Map
-
-
Change BookStoreMapper
@Component class BookStoreMapper: EntityMapper<BookStore, UUID>() { override fun EntityTypeDSL<BookStore, UUID>.config() { ... other configuration ... userImplementation(BookStore::avgPrice) // α } fun avgPrice() = // β runtime.batchImplement(BookStore::avgPrice) { // γ spring(BookRepository::class) // δ .findAvgPriceGroupByStoreIds(it) // ε } }
-
α
Map BookStore.avgPrice as user implementation field
-
β
Provides a public function.
-
In theory, the function name is arbitrary, but for readability it is recommended to keep the same as the name of BookStore::avgPrice at γ
-
This function has no return type
-
-
γ
-
runtime is a protected property provided by the superclass org.babyfish.graphql.provider.EntityMapper
-
BookStore::avgPrice represents the field for which you want to implement by your code
-
Here, the function we call is not runtime.implementation, but runtime.batchImplementation. This means that BookStore.avgPrice can be optimized by DataLoader
-
-
δ
spring() is a protected function provided by the superclass org.babyfish.graphql.provider.EntityMapper
In Spring dependencies, the dependency chain is as follows:
BookRepository ➤ R2dbcClient ➤ SqlClient ➤ All Entity MappersIf we directly inject BookRepository into BookStoreMapper here, spring will throw an exception because of the circular dependencies problem.
The solution provided by Spring for this scenario is @org.springframework.beans.factory.annotation.Lookup.
To simplify this problem, org.babyfish.graphql.provider.EntityMapper provides the spring() function, which can help us get external dependencies that cannot be injected directly.
-
ε
it is an implicit variable provided by the lambda expression of the runtime.batchImplementation function, it represents the collection formed by the ids of all BookStores in a DataLoader batch processing.
The length of this collection is controlled by the batchSize configuration of the BookStore.avgPrice, We have already discussed the bachSize configuration in Configure batch size, so I won't repeat it here.
-
Start app, access http://localhost:8080/graphiql and execute
query {
books {
name
store {
name
avgPrice
}
}
}
The response is
{
"data": {
"books": [
{
"name": "Effective TypeScript",
"store": {
"name": "O'REILLY",
"avgPrice": 56.833333333333336
}
},
{
"name": "GraphQL in Action",
"store": {
"name": "MANNING",
"avgPrice": 80
}
},
{
"name": "Learning GraphQL",
"store": {
"name": "O'REILLY",
"avgPrice": 56.833333333333336
}
},
{
"name": "Programming TypeScript",
"store": {
"name": "O'REILLY",
"avgPrice": 56.833333333333336
}
}
]
}
}
< Previous: Add arguments to association | Home | Next: Pagination query >