Introspection Queries With GraphQL

With GraphQL you can run introspection queries to learn about the available fields and types of a GraphQL schema. That introspection capability is also what gives GraphiQL the ability to provide documentation about the schema and autocompletion.

Let’s use the Star Wars SWAPI public API endpoint to run a few introspection queries against.

This affiliate banner helps support the site 🙏

__type

First let’s run a query to enquire about the Film type using the built-in __type:

query FilmType {
  __type(name: "Film") {
    kind
    name
    fields {
      name
      description
      type {
        name
      }
    }
  }
}

• Kind gives us the enum value for the type, like OBJECT, SCALAR or INTERFACE.
• Name gives us the name of the type.
• Description, well, gives us the description!

And here’s what the response looks like:

{
  "data": {
    "__type": {
      "kind": "OBJECT",
      "name": "Film",
      "fields": [
        {
          "name": "title",
          "description": "The title of this film.",
          "type": {
            "name": "String"
          }
        },
        {
          "name": "episodeID",
          "description": "The episode number of this film.",
          "type": {
            "name": "Int"
          }
        },
        ...

Notice the use of the built-in __type (of type __Type) here to get information on the type of a particular object or interface.

Here’s another example using a fragment to learn even more about a particular type:

query LearnAboutFilm {
  __type(name: "Film") {
    ...AboutType
  }
}

fragment AboutType on __Type {
  fields {
    name
    description
    args {
      name
      description
    }
  }
  interfaces {
    name
    description
  }
  inputFields {
    name
    description
  }
  possibleTypes {
    kind
    name
    fields {
      name
      description
      type {
        kind
        name
        description
      }
    }
  }
}

__schema

With __schema we can ask the server about the schema itself. Let’s look at an example:

query LearnAboutSchema {
  __schema {
    types {
      name
      kind
    }
    queryType {
      fields {
        name
        description
      }
    }
  }
}

And the response:

{
  "data": {
    "__schema": {
      "types": [
        {
          "name": "Root",
          "kind": "OBJECT"
        },
        {
          "name": "String",
          "kind": "SCALAR"
        },
        {
          "name": "Int",
          "kind": "SCALAR"
        },
        ...
        "queryType": {
        "fields": [
          {
            "name": "allFilms",
            "description": null
          },
          {
            "name": "film",
            "description": null
          },
          ...

__typename

__typename can be used as part of regular queries to enquire about the type of a particular field:

query LearnAboutFilm {
  allFilms {
    films {
      __typename
      title
    }
  }
  film (id: "ZmlsbXM6Mw==") {
    __typename
    title
  }
  starship(id: "c3RhcnNoaXBzOjc1") {
    __typename
    name
    model
  }
}

And here’s the response:

{
  "data": {
    "allFilms": {
      "films": [
        {
          "__typename": "Film",
          "title": "A New Hope"
        },
        {
          "__typename": "Film",
          "title": "The Empire Strikes Back"
        },
        ...
      ]
    },
    "film": {
      "__typename": "Film",
      "title": "Return of the Jedi"
    },
    "starship": {
      "__typename": "Starship",
      "name": "V-wing",
      "model": "Alpha-3 Nimbus-class V-wing starfighter"
    }
  }
}