About design schema for E-commerce database

Users Dgraph
,
1

in the relational database, it is connected by ID:
Table->User: u_ id ,u_ name
Table->Product : p_ id,p_ tittle

Table->Trade: t_ id,u_ id (buyer), p_id (Product ID)

How to design it in dgraph Schema?

2 
type User {
  id: String! @id
  name: String
}

type Product {
  id: String! @id
  title: String
}

type Trade {
  id: String! @id
  buyer: User!
  bought: Product!
}

This will allow you to import data, but this is not preferred for an empty system. For that, I would use the ID scalar on the id properties instead of the String! @id. You mentioned existing database so I suppose you need to import those ids.

1 Like
3

I’d like to take this opportunity to ask @amaster507 on how one would shift mindset from a table-based SQL structure to graph based structure. If you could share, I could split and make it a wiki post.

1 Like
4

In graphDBs you don’t have to care about foreign keys, join tables and so on. Graphs are all about relations.

Your schema would look like

PS: This is a DQL Schema, not GraphQL.

type User {
  name
  phone
}

type Product {
  title
  barcode
}

type Trade {
  tax
  from
  products
}

  name: string @index(exact) .
  phone: string .
  title: string @index(exact) .
  barcode: string @index(exact) .
  tax: int .
  from: uid @reverse .
  products: uid @reverse .

Your mutation

{
   "set": [
      {
         "tax": "0%",
         "from": {
            "uid": "0xa31"
         },
         "products": [
            {
               "uid": "0x21"
            },
            {
               "uid": "0x22"
            },
            {
               "uid": "0x23"
            }
         ]
      }
   ]
}

And you can do calculations on the fly with queries.

2 Likes
5

@amaster507
@MichelDiz
Thank you for your answers and efforts.
The question was answered.