Field
module Types
class QueryType < GraphQL::Schema::Object
field :restaurants, [RestaurantType], null: true do
description "Lists all restaurants"
end
def restaurants
Restaurant.all
end
end
end
Query
query Restaurants {
restaurants {
name
cuisine
}
}
In the example above, Restaurants.all will initially get all the rows from the restaurant table and then filter them out according to the query.
This is quite inefficient as we can use the query to filter out our data on the actual sql request:
module Types
class QueryType < GraphQL::Schema::Object
field :restaurant, RestaurantType, null: true do
description "Find a restaurant by ID"
argument :id, Int, required: true
end
field :restaurants, [RestaurantType], null: true, resolve: -> (obj, args, ctx) {
children = ctx.irep_node.scoped_children
root_key = children.keys.first
params = children[root_key].keys
Restaurant.all.select(params.join(", "))
} do
description "Lists all restaurants"
end
end
end
The underline sql representation is then
Option A
SELECT "restaurants".* FROM "restaurants" - returns more rows than needed
Option B
SELECT name, cuisine FROM "restaurants" - only returns the requested rows
This is not ideal as it won't work with joins, unless explicitly handled.
Since I'm quite new to Ruby and/or Rails I've been wondering if that's a good practice, or are there any other ways to achieve such optimization?
