Exercise: Using dispatchers

For each of the functions below, decide if you need to set a dispatcher, and if so, which one:

• createInvoice - prepares an invoice and sends it to the SaaS API using the Retrofit library. In this library, we define suspending functions. The toFakturowniaInvoiceData function is a simple mapping.
• sendEmail - prepares an email and sends it using the api method of the SendGrid class. This is a blocking operation that returns a Response object.
• getUserOrders - gets orders made by the user from the database using the MongoDB Kotlin driver. toList is a suspending function that returns a list of orders.
• upscaleImage - a function that uses the TensorFlow library to upscale an image.

override suspend fun createInvoice(
    invoiceData: InvoiceData,
) {
    invoiceApi.postInvoice(
        invoiceData.toFakturowniaInvoiceData()
    )
}

interface InvoiceApi {
    @POST("invoices.json")
    suspend fun postInvoice(
        @Body invoice: SendInvoiceData
    ): InvoiceCreationResponse
}

private val invoiceApi = retrofit2.Retrofit.Builder()
    .baseUrl(invoiceBaseUrl)
    .build()
    .create(InvoiceApi::class.java)

private val sendGrid = SendGrid(API_KEY)

suspend fun sendEmail(email: Email) {
    val request = Request().apply {
        method = Method.POST
        endpoint = "mail/send"
        body = email.toSendGridEmail()
    }
    sendGrid.api(request)
}

private val orderCollection = mongoClient
    .getDatabase("shop")
    .getCollection<Order>("orders")

suspend fun getUserOrders(userId: String): List<Order> =
    orderCollection
        .find(eq("userId", userId))
        .toList()

val model = TensorFlowModel()

suspend fun upscaleImage(image: Image): Image =
    model.upscale(image)

Once you are done with the exercise, you can check your solution here.