hey guys! how do you design the resolvers so that it wont clutter so much? Like i m using the howtographql example (

resolvers/Query.js

,

resolvers/Mutation.js

)... can I divide them into let's say

User/Query.js

and

User/Mutation.js

and other models following the same design?

It's just Javascript files, so as long as you require them correctly, you can divide them any way you want.

For bigger projects, I think having a

resolvers

folder with

queries

and

mutations

as subfolders is also useful. Alternatively, you can have queries and mutations grouped together by the types. As in

my_app/src/users/queries.js

,

my_app/src/users/mutations.js

In my projects, I have a

resolvers

folder with a few files (user.js, stuff.js, ...). In each of these file, I do the following:

export {
  Query: {
    myQuery1: ...
  },
  Mutation: {
    myMutation1: ...
  }
}

Which is better to group the logic I think

@Errorname : What do you do if a mutation connects a

user

to an object (say

post

or

comment

)?

I'm not sure I understand your question. Are you talking about a user mutation which calls a post mutation ?

The last one

Well, it depends. First, I would do the connection directly in the

createComment

mutation. But if you specifically want a mutation which only connect the two, I would probably put it in comments.js as User is kind of a "super-object" (used nearly everywhere). Let's imagine you wanted to completely remove the comment feature of your app: You would only need to delete the

comments.js

file!

I think it’s a good solution. I was just curious how you would solve that one

the way i m using is having

resolvers/Customer/Customer.js

as the class, then

resolvers/Customer/Mutation.js

as the mutation and

resolvers/Customer/Query,js

as the queries... so per model per directory