On Thu, Jun 30, 2016 at 7:56 PM, Nicholas Alexander <nalexan...@mozilla.com> wrote: > Hi Gabriel, > > The "whole code need[s] to be rewritten", but it's really not as bad as it > looks. The WebChannel stuff is just some browser-level abstraction around a > simple event passing model, and in fact you can opt out of it entirely (at > least for a while). There's an example of doing the dance in a WebView in > the iOS repository. You can start digging into the code at > > https://github.com/mozilla/firefox-ios/blob/5238e873e77e9ad3e699f926d12f61ccafabdc11/Account/FirefoxAccountConfiguration.swift#L95 > > and > > https://github.com/mozilla/firefox-ios/blob/5238e873e77e9ad3e699f926d12f61ccafabdc11/FxAClient/Frontend/SignIn/FxASignIn.js > > and > > https://github.com/mozilla/firefox-ios/blob/e8f885d0f169f55dd466838a09ba0a440b6bc83c/FxAClient/Frontend/SignIn/FxAContentViewController.swift#L20. > > The first link includes `service=sync&context=fx_ios_v1`; that'll ensure the > accounts service sends you the messages over postMessage. The second link > shows how to inject the message listener into the WebView context, so you > can react to the messages. The third link shows what we do with those > messages -- the important one is "login", which includes all the tokens > you'll need to talk to the auth server to get keys, etc. > > This postMessage interface has been cleaned up and regularized into the > WebChannel interface, but the underlying mechanism is almost identical. The > browser bits are just an abstraction on top of that since we want to use the > same approach for many browser <-> content interactions.
Thanks Nick, I'll give it a look tomorrow, now it is really late here. Regards, Gabriel _______________________________________________ Sync-dev mailing list Sync-dev@mozilla.org https://mail.mozilla.org/listinfo/sync-dev
