Add support for new passwordless endpoint

[nbandarchi]

Changes

Fixing PasswordlessAuthenticator.signIn()

• Changed grant type to http://auth0.com/oauth/grant-type/passwordless/otp
• Mapping connection to realm and password to otp
• Changing default type from ro to token in OAuthAuthenticator.signIn()
• Allowing type to be specified in options

References

Passwordless signin process is attempting to use oauth/ro endpoint which no longer exists:

• https://support.auth0.com/tickets/00472740

Testing

• This change adds unit test coverage
• This change adds integration test coverage

Checklist

• I have read the Auth0 general contribution guidelines
• I have read the Auth0 Code of Conduct
• All existing and new tests complete without errors

[davidpatrick]

@nbandarchi thanks for raising the PR and addressing this issue. Unfortunately oauth/ro does still exist when a customer has an allow legacy grant type flag enabled on their account. So in order to accept this PR without cutting a major we will need keep the default behavior of the passwordless signIn function.

We can add a deprecation notice to that function and that it will move to legacySignIn on the next major. And in order to handle the functionality of the new endpoint, we can add the otp and realm parameter and when either is detected switch the other parameters to match the new endpoint.

[nbandarchi]

Updated so that it's not defaulting to token. if passwordless.signIn() is called with otp and optionally realm it will change the grant type to http://auth0.com/oauth/grant-type/passwordless/otp and pass in token as an option to oauth.signIn(). Realm will default to sms the same way that connection currently does.

[davidpatrick]

@nbandarchi thanks for the quick turnaround! Can you please document the otp and realm parameters in the jsdoc area for that function? Also we'll need some new tests in passwordless.tests.js for #signIn to cover the new otp usage.

Thanks 🙏

[nbandarchi]

No problem! Added tests and documentation. The tests are just a duplicated and updated copy of the current tests, but still should show that both paths can coexist.

[davidpatrick]

Changes are great! Thanks for the work here