Value required for key 'access_token'.

[PanArtemG]

Hey! Can you please tell me what is the problem?
I receive a response from FB and Google with the code but I cannot process it. The execution process is interrupted before reaching the last argument "completion" to process the incoming token and then get user data
try routes.oAuth(from: Facebook.self, authenticate: "login-facebook", callback: facebookCallbackURL, scope: ["public_profile", "email"], completion: processFacebookLogin)

[PanArtemG]

If I add another route with the same address as facebookCallbackURL, the application swears that the route is being overwritten Warning: Overriding route output at: GET/oauth/facebook, but then a code comes to me and I can process it

`try routes.oAuth(from: Facebook.self, authenticate: "login-facebook", callback: facebookCallbackURL, scope: ["public_profile", "email"], completion: processFacebookLogin)

routes.get("oauth", "facebook", use: getCode)`

Code received

[PanArtemG]

SigIn with Google creates the same error

[PanArtemG]

Hi!
The problem was my own, global JSON decoding was connected in the project
`let encoder = JSONEncoder()
encoder.keyEncodingStrategy = .convertToSnakeCase
encoder.dateEncodingStrategy = .iso8601
ContentConfiguration.global.use(encoder: encoder, for: .json)

let decoder = JSONDecoder()
decoder.keyDecodingStrategy = .convertFromSnakeCase
decoder.dateDecodingStrategy = .iso8601
ContentConfiguration.global.use(decoder: decoder, for: .json)`

[0xTim]

Reopening as Imperial should handle this better regardless of what global settings people have

[froggomad]

Interestingly, I'm getting a 200 back, but it comes with an error:

error=incorrect_client_credentials&error_description=The+client_id+and%2For+client_secret+passed+are+incorrect.&error_uri=https%3A%2F%2Fdocs.github.com%2Fapps%2Fmanaging-oauth-apps%2Ftroubleshooting-oauth-app-access-token-request-errors%2F%23incorrect-client-credentials

• status : NIOHTTP1.HTTPResponseStatus.ok
  ▿ headers : Server: GitHub.com

I'm probably using the wrong client secret or id judging by the URI I get back...

Edit: Yep, wrong client secret in my case

[j4nos]

I have almost the same issue.

Value of type 'String' required for key 'access_token'.

Something wrong here?

let githubMediaType = HTTPMediaType(type: "application", subType: "vnd.github+json")
ContentConfiguration.global.use(decoder: JSONDecoder(), for: githubMediaType)
ContentConfiguration.global.use(encoder: JSONEncoder(), for: githubMediaType)

Based on Oauth 2.0 spec access token is not needed in step 5, but it is fetched and returned in step 7.

[danpalmer]

Hi all, just wanted to say that I've unfortunately just lost an hour of debugging and stepping through a bunch of Imperial code to find this bug.

I think it's a clear bug that encoders/decoders customised by the Vapor installation are being used for interaction with upstream APIs that have an entirely different set of compatibility requirements.

It seems this affects at least Google and Facebook, and quite possibly others.

This will probably need an extension to each of the providers for them to define their own content decoder, and then use that instead of the global one that is used implicitly by https://github.com/vapor/vapor/blob/main/Sources/Vapor/Content/ContentContainer.swift#L128.

[talmeme]

I get the same error as in the issue subject with Keycloak. Making the following change worked for me:

--- a/Sources/ImperialKeycloak/KeycloakRouter.swift
+++ b/Sources/ImperialKeycloak/KeycloakRouter.swift
@@ -9,6 +9,7 @@ public class KeycloakRouter: FederatedServiceRouter {
     public let callbackURL: String
     public let accessTokenURL: String
     public let service: OAuthService = .keycloak
+    public let callbackHeaders = HTTPHeaders([("Content-Type", "application/x-www-form-urlencoded")])
     
     public required init(callback: String, completion: @escaping (Request, String) throws -> (EventLoopFut>
         self.tokens = try KeycloakAuth()

[bbatth]

@0xTim, I dont see this above change for keycloak merged in main branch, I had the same issue and fixed by making this change locally.

[0xTim]

@bbatth I don't think a patch was ever raised