SwiftData within Observable class

Community
1

Greetings Everyone. I have a question regarding my capstone project. I have an app (which I will introduce later in my Journal) that has more than one Swift Data model. Furthermore, I currently have all the models loaded in one class with observation and distributed in every view that is needed with the @environment macro. Although until now, the app worked smoothly, but I have second thoughts regarding if this is the correct way of using swift Data models in an app.
Thanks

2

@Yianni1980
Welcome to the community.

Without seeing your code I am not sure what you mean by having all the models in the one class. That said, if it works for you as is then great.

3

I am sorry, you are right. This is my diabetes management app. This class logs glucose readings, meals, medications. Until now, I used @Query and Model Context in every view needed. Using centralized all in one class and I can say it works as expected, but I don’t know if it is the correct approach. Also, ultimately should I use SwiftData or move to remote (supabase,firebase)?

@Observable
class HealthDataStore {
var glucoseReadings: [GlucoseReading] =
var meals: [Meal] =
var medications: [MedicationLog] =
let context: ModelContext

init(context:ModelContext) {
    self.context = context
}



var latestGlucoseReading: Double {
    let todayReadings =
    glucoseReadings.filter {
        Calendar.current.isDateInToday($0.timestamp) }
    return todayReadings.isEmpty ? 0 : todayReadings.first!.value
}

var sevenDayAverageGlucose: Double {
    averageGlucose(inLastDays: 7)
}

var latestReadingTimeAgo: String {
    guard let latest = glucoseReadings.sorted(by: { $0.timestamp > $1.timestamp }).first else {
        return "No readings"
    }

    let formatter = RelativeDateTimeFormatter()
    formatter.unitsStyle = .full
    return formatter.localizedString(for: latest.timestamp, relativeTo: Date())
}


func fetchAll() {
    do {
        print("fetching from database")
        glucoseReadings = try context.fetch(FetchDescriptor<GlucoseReading>(sortBy: [.init(\GlucoseReading.timestamp, order: .reverse)]))
        meals = try context.fetch(FetchDescriptor<Meal>(sortBy: [.init(\Meal.timestamp, order: .reverse)]))
        medications = try context.fetch(FetchDescriptor<MedicationLog>(sortBy: [.init(\MedicationLog.timestamp, order: .reverse)]))
        print("finishe fetching from database")

    } catch {
        print("Fetch failed:", error)
    }
}

func addGlucose(_ reading: GlucoseReading) {
    context.insert(reading)
    try? context.save()
    fetchAll()
}

func addMeal(_ meal: Meal) {
    context.insert(meal)
    try? context.save()
    fetchAll()
}

func addMedication(_ med: MedicationLog) {
    context.insert(med)
    try? context.save()
    fetchAll()
}

}

4

OK I understand what you are doing there. The way I have learned MVVM is such that I would have called that a ViewModel which is your Observable class. I would assume that each of GlucoseReading, Meal and MedicationLog are your individual models defined as structs and declared either in individual files or collectively in a single file with each one having @Model preceding the struct definition and some @Relationship to link them together where necessary.

What you are doing with that Observable class is much the same as I would have done it.

5

Each Model (glucose reading,meal,medicationLog) are classes with @Model declared. and they don’t have a relationship with each other. Assuming that a user can log everything he wants without the need to be linked the models together. If you want i can put some code regarding the declaration of the models for each measurement individually without any relationship with each other. What do you think?

6

No that’s not necessary to post your model declarations code. If there is no relationship with any of the models then that makes it easier for you.