Identy.Normalizer

normalize :: forall a. ReadForeign a => Foreign -> E a

Normalize nested entity to identy-style, then decode normalized foreign.

When should you use this ?

API in the world often respond nested-style object like this:

{
  "id": "1",
  "name": "User Name",
  "comments": [
    { "id": "2", "body": "Comment Body 2" },
    { "id": "1", "body": "Comment Body 1" }
  ]
}

normalize formats object from nested-style to identy-style.

{
  "entities": {
    "user": { "1": { "id": "1", "name": "User Name" } },
    "comment": { "1": { "id": "1", "body": "Comment Body 1" }, "2": { "id": "2", "body": "Comment Body 2" } }
  },
  "associations": {
    "userComments": { "1": [ "2", "1" ] } // from parent user id to comment ids.
  },
  "result": "1" // This is root entity id. In this case, user id.
}

result has root entity id. Therefore, if a response is single entity, result is an id of its entity, if a response is array of entity, result is ids of entities.

Prerequisite:

Each entity need a property typename that is set type name with upper camel case.

For example:

{
  "id": "1",
  "name": "User Name",
  "typename": "User"
  "comments": [
    { "id": "2", "body": "Comment Body 2", "typename": "Comment" },
    { "id": "1", "body": "Comment Body 1", "typename": "Comment" }
  ]
}

Note:

Essentially, such formatting for clients is the work of API server.

If it's possible, receive identy-style JSON directly from API server.