Batchloaders nesting

[MarianneHorsch]

Hi all,

I'm struggeling with my Graphql service to get my scenario working. Lets asume I have this schema:

type Query {
  companies: [Company]
}

type Company {
  companyId: Int
  departments: [Department]
}

type Department {
  departmentId: Int
  testers: [Employee]
  softwareEngineers[Employee]
}

type Employee {
  employeeId: [Int]
  name: String
}

My companies come from the CompanyService, departments can be retrieved in batches based on the companyID and return a list with employeeIds that are testers and a list with employeeIds that are softwareEngineers, employers can be retrieved in batches from the EmployeeService based on employeeId.
So first I created a MappedBatchLoaderWithContext (context is only because of my real case needs context) to fetch the departments then when the results are in I want to have another MappedBatchLoaderWithContext to fetch the employers. I used. a DepartmentsDataFetcher and a EmployerDataFetcher but no matter what I do it seems that the graphQL query returns its result before the EmployeeService data is added. I see the EmployeeService is being called, I see the response in the dataloader but it never ends up in my query.
I read the part about Chaining Data loaders but since my EmployeeService only returns an employee and nothing about the department the code is not pretty (I need to merge the 2 objects myself) I would prefer to get this working with 2 seperate datafetchers one for type Department and one for type Employee, this makes the Employee datafetcher a third step in execution (first the companies, second in batches the deparments and third iteration the employers). Is this possible with DGS?? How can I solve this issue?
Thank you!

[congotej]

Hi @MarianneHorsch! Would you mind sharing your datafetcher/dataloader code so I can better understand how you are implementing this?

It's a bit convoluted to understand exactly what you have currently, but sounds like you are on the right track.
You could have a dataloader for departments, and another for employees and then use those in your respective datafetchers.
Something like :

@DgsDataLoader(name = "departments")
public class DepartmentLoader ... {
}

@DgsDataLoader(name = "employees") 
public class EmployeeLoader ... {
}

@DgsData(parentType = "Company", field = "departments")
// call departmentDataLoader

@DgsData(parentType = "Department", field = "employees")
// call employeeDataLoader

Again, this is just hypothetical, but I can provide more pointed feedback if you share your implemented logic.