[corona]

Do you have a specific site in mind? I couldn't find one.

I think with some effort one can figure out the layout of their data to a sufficient degree to get all of the information needed by the plugin. I've got most of that, but not all. It's a matter of monitoring the requests the pages make and then some guesswork on top of that. You'd probably still need the main web site for authentication. I wonder what the lifetime of their authentication tokens is.

@hstegeman and I were discussing this a good few pages back: http://forum.kodi.tv/showthread.php?tid=...pid2092521
http://beta.istreamguide.com was the best one for having some nice clean json, but they've since closed the beta site for some closed dev/testing/planning :-(

I like yidio.com for having a good json api again ( http://cf.yidio.com/ajax_base_dir_shows_...letter=All )
but it's unclear how well they support international netflix sites.

It's also quite different in that it's an aggregater of streaming services. It can be set to just display shows from netflix, or it can display shows from heaps of different services and then play them from wherever.

From my initial testing (in australia) when set to filter for netflix it only seems to show things available in your country, but then doesn't give me a netflix link to watch them. I've sent them a question about their international support.

The pages that the netflix website serves contain pointers to the REST servers. They are on the netflix.com domain, which is hardcoded for no good reason in my code. The rest of the URL comes from the page though. It seems that you need that and an authentication token from the main page.

Yeah it's possible for third party sites to permanently link to a netflix account, yidio does this, as such it might be worth folling the network requests it uses when you go through the linking process to see if there's any netflix api exposed we could use to achieve the same goal (I doubt it),

Otherwise we do still need to script this part on the netflix site. My new webdriver module leverages selenium which means you can automate the button clicks/textbox form submissions more easily if the api endpoint can't be found. Then the cookies/session gets stored to keep logged in as long as possible (already done).