Hi Dmitry,
On 04/04/2019 00:45, Dmitry Telegin wrote:
Hi Marek,
You absolutely right, UsernamePasswordForm does the trick. However, the login screen
rendered by UsernamePasswordForm is different from that of IdpUsernamePasswordForm in the
following aspects:
- IdpUsernamePasswordForm doesn't display the block with IdP/social buttons
You're right. Small addition: The IdpUsernamePasswordForm displays
social buttons, but just of those identity providers, which are already
linked to specified user. In other words, if you want to link your
account to broker-A and your account is already linked to broker-B, then
broker-B is displayed on the form. This way, you have possibility to
re-authenticate not just with your password, but alternatively by login
to already linked broker-B, which is already linked to your account and
hence "trusted" to be used for prove your identity.
It seems that with your proposal in case that username is unknown, we
won't display any brokers on the screen and hence it will be mandatory
to do re-authentication by username+password?
- IdpUsernamePasswordForm renders the message relevant to
IdP-linking-by-reauthentication, which is this:
federatedIdentityConfirmReauthenticateMessage=Authenticate as {0} to link your account
with {1}
So, my requirement is to implement the appearance of IdpUsernamePasswordForm + behavior
of UsernamePasswordForm. I think this could be done either by augmenting the former, or by
merging the two authenticators into a unified one, that would exhibit different behavior
depending on the context (normal login vs. reauthentication for IdP linking).
I suggest to update IdpUsernamePasswordForm authenticator. In case that
EXISTING_USER_INFO is not there, we can do the behaviour like:
- User will need to provide both username+password. Hence username field
will need to be enabled
- Social buttons won't be displayed on the login screen
- Message will be bit different. For example just: Authenticate to link
your account with {1}
For the case when EXISTING_USER_INFO is available, I would like to keep
the same behaviour as currently is.
WDYT?
Marek
Please let me know which way seems better for you, with the idea in mind of having this
contributed to upstream.
Thanks!
Dmitry
On Tue, 2019-04-02 at 15:21 +0200, Marek Posolda wrote:
> On 28/03/2019 17:06, Dmitry Telegin wrote:
>> Hi,
>>
>> I'm currently working to implement the following requirements:
>> - users are managed externally via LDAP, self-registrations disabled;
>> - there is an external IdP;
>> - generally, there is no way to automatically match IdP identity with
Keycloak's one, so IdP linking will always be performed by the user manually;
>> - in order to do that, the user should click the IdP icon in the login screen,
authenticate with the IdP, get back to Keycloak and "claim" his/her Keycloak
account by entering correct username and password.
>>
>> Currently, the closest thing in Keycloak is
o.k.authentication.authenticators.broker.IdpUsernamePasswordForm (aka
"idp-username-password-form", aka "Username Password Form for identity
provider reauthentication").
>> However, it 1) prefills username field and makes it non-editable, 2) depends on
the preceding IdpCreateUserIfUniqueAuthenticator execution to provide existing user model
(EXISTING_USER_INFO auth note).
>>
>> My proposal is to improve IdpUsernamePasswordForm by allowing its execution even
without the preceding IdpCreateUserIfUniqueAuthenticator. In the absence of
EXISTING_USER_INFO, IdpUsernamePasswordForm should allow the user to manually enter
username.
> I wonder if you can't already achieve something like this with the OOTB
> authenticator implementations, but just correctly configure them? For
> example in the "First Broker Login" flow used for your identity
> provider, you can just directly use the default browser-based
> authenticator ( UsernamePasswordForm ) instead of the
> IdpUsernamePasswordForm. That way, the username+password form will be
> always shown for "First Broker Login" and once user authenticates, his
> account will be linked with IdP account.
>
> Marek
>
>> Please let me know if you think it's worth having this in Keycloak. Regards,
>> Dmitry
>>
>> _______________________________________________
>> keycloak-dev mailing list
>> keycloak-dev(a)lists.jboss.org
>>
https://lists.jboss.org/mailman/listinfo/keycloak-dev
>