0

I'm currently working on refactoring a project to use the repository pattern, but I'm currently struggling with how related information should be retrieved.

Let's say I have a Hotels and Rooms. I have 3 use-cases that come up:

  • Retrieve a hotel with all of its rooms
  • Retrieve a hotel (without rooms)
  • Retrieve a room

Now the last 2 seem simple enough to me, I would have some kind of findHotelById and findRoomById methods in my repositories. Now what is the best practice for retrieving both the hotel and the rooms? Should I have a findHotelByIdWithRooms method? Should I call findHotelById and then call findRoomsByHotel and populate the rooms on my model (ex: hotel.setRooms(rooms)) after the fact? Or should findHotelById take in some parameter to optionally include rooms in the result?

Matthew Weeks
  • 141
  • 2
    Can this answer your question? Repository Pattern and Joined Queries – Fabio Oct 06 '21 at 18:55
  • I'm not totally sure. I see in that example they have a GetWithLinesById method, which would appear to be the same as my FindHotelByIdWithRooms example. Although I don't feel that I am attempting to mirror the database like mentioned in that issue, as Rooms and Hotels would be 2 separate entities – Matthew Weeks Oct 06 '21 at 19:12
  • 1
    Upon second glance, I think you're right. The main part to focus on being So, instead of mirroring database structure in your repository, have abstractions which fit domain needs and then implement those abstractions by effectively using database features. – Matthew Weeks Oct 06 '21 at 19:23
  • To me this question is an indication of a model that hasn't been fully fleshed out. How is it that that a Hotel could both have and not have a collection of Room? What I'm saying is a "hotel with rooms" is a different entity than a "hotel without rooms". Your confusion is a result of trying to bridge this gap without updating your understanding of the model. – user3347715 Oct 09 '21 at 15:24

0 Answers0