Clojurians Log v2

can a pathom mutation utilize resolvers to load data, or can it only utilize parameters explicitly passed to it? for example, if I pass a user/id to the mutation, can i use my already written resolves to load other fields, or would i have to load them from explicitly from whatever data-source I’m using?

@tylernisonoff you can do

(pc/defmutation do-thing [{:keys [parser]
                           :as   env} {:user/keys [id]}]
  {::pc/params [:user/id]}
  (let [eid [:user/id id]
        user-with-address (-> env
                              (parser [{eid [:user/id
                                             :user/address]}])
                              (get eid))]
    ...))

awesome! trying that now

perfect — thanks @souenzzo

I use this small helper in my codebase. I don't like the name of this function but it's useful.

(defn get-from-context
  [{:keys [parser] :as env} context query]
  (let [[eid & others] context]
    (-> (parser env `[{(~eid {:pathom/context ~(into {} others)})
                       ~query}])
        (get eid))))
(pc/defmutation do-thing [env params]
  {::pc/params [:user/id]}
  (let [user-with-address (get-from-context env params [:user/id
                                                        :user/address])]
    ...))

this is possible to implement with a plugin, the gist of it is to get the source params and run a query using that as the initial context

you can do a recursive call to the parser (available in the env), setting ::p/entity to be a (atom params), then send the response of that to the mutation, makes sense?

ups, just realized @souenzzo already replied :)

good to know about the plugin idea too 🙂 thanks!

as a suggestion, creating a custom ::pc/transform can make this look nicer

plugin, or a function to be used as ::pc/transform

that way you can cherry pick which mutations should "auto-resolve" params