add postgres support #32
Conversation
| amoa_provider VARCHAR(255) NOT NULL, | ||
|
|
||
| -- the local user id | ||
| amoa_local_user INTEGER NOT NULL, |
There was a problem hiding this comment.
I could only identify this to have changed from INTEGER UNSIGNED to INTEGER. Do you by chance know why it was INTEGER UNSIGNED for mysql and why we could not simply make it INTEGER for both? I would like to keep this as simple as possible.
There was a problem hiding this comment.
I believe I fixed this with
| @@ -240,7 +240,7 @@ public function continuePrimaryAuthentication( array $reqs ) { | |||
| $result = $dbr->select( | |||
| 'authmanageroauth_linked_accounts', | |||
| [ 'amoa_provider', 'amoa_remote_user', 'amoa_local_user' ], | |||
| [ 'amoa_provider' => $req->amoa_provider, 'amoa_remote_user' => $resp->linkRequest->amoa_remote_user ], | |||
| [ 'amoa_provider' => $req->amoa_provider, 'amoa_remote_user' => strval($resp->linkRequest->amoa_remote_user) ], | |||
There was a problem hiding this comment.
For which cases does this make a difference? I currently don't have the big picture in my head.
There was a problem hiding this comment.
Tested myself, seems like amoa_remote_user can be an integer if the identity provider provides it in that format.
There was a problem hiding this comment.
I believe I fixed this with
|
Thanks for providing pointers at what needed fixing. As explained in my comments I think I fixed this in my yearly cleanup 4dfbb9d (with a few different changes that hopefully lead to the same result). Also sorry for taking so long, I missed the notification. Feel free to ping or request review in that case next time. While my changes in the end are quite different to yours I should've probably still mentioned you somewhere in the fix, sorry. |
No description provided.