* feat(screenshare): emit source type when starting screenshare

* squash: update doc

Bring over two fixes for spot. One is for
identifying the screenshare type when using
a camera for screenshare or when using a proxy
stream. Also bring in a fix to avoid a js error
in chrome ios.

* Adds integrations doc.

Google, Microsoft and Dropbox for now.

* Updates doc.

Seems like windows does not have emojis for flags.

* Adds new format of phoneList service and re-design dial in numbers page.

Adds flags and country names (with translations) for the numbers if using the new format.

* Fixes tests and fixes get default number.

* Updates swagger with new format.

* Moves html back yo table.

Fixes displaying on mobile and also the tel: URI generation. The tel: URI is tested on Android and iOS and seems to work (Android was not interpreting 'p', but both seems to like ',').

* Fixes a wrong return statement.

* Small fixes.

Update api.md

When a conference is started, the currently used
camera device id is saved. I believe this is happening
because lib-jitsi-meet does not use exact device id
mathcing when calling getUserMedia, so it's possibl
to request camera A but get camera B back because
camera A is not available. When config.startScreenSharing
is true, the syncing occurs and saves the desktop
source id. So when screensharing is stopped, jitsi-meet
requests that desktop source id instead of the preferred
camera.

Adds a new event fired when Jitsi Meet shows the feedback prompt.

The conference disconnection process is asynchronous which means there's
no guarantee that there will be CONFERENCE_LEFT event for the old
conference, before the next conference is joined. Because of that we can
end up with two simultaneous calls on the native side which is not
always supported. End the call on CONFERENCE_WILL_LEAVE to fix this
corner case.

Using an array of styles in Firefox causes an error
that triggers jitsi-meet to redirect to a static page.

They generate a bunch of harmless yet confusing error messages and they are not
really useful.

RTCAudioSession is a thin wrapper around AVAudioSession provided by the WebRTC
framework. It makes some use-cases easier, and leads us closer to manual audio
unit management, which we will likely need in the near future.

If a new conference is joined and as a result the current one is terminated,
don't ever attempt to set the audio mode.

Fixes this issue:

Main Thread Checker: UI API called on a background thread: -[UIApplication setExpectsFaceContact:inLandscape:]
PID: 25442, TID: 10886619, Thread name: (none), Queue name: com.facebook.react.ProximityQueue, QoS: 0
Backtrace:
4   JitsiMeet                           0x000000010a0eaadc -[Proximity setEnabled:] + 64
5   CoreFoundation                      0x00000001fea34630 <redacted> + 144
6   CoreFoundation                      0x00000001fe912450 <redacted> + 292
7   CoreFoundation                      0x00000001fe913034 <redacted> + 60
8   JitsiMeet                           0x000000010a4e08e8 -[RCTModuleMethod invokeWithBridge:module:arguments:] + 492
9   JitsiMeet                           0x000000010a4e7a10 _ZN8facebook5reactL11invokeInnerEP9RCTBridgeP13RCTModuleDatajRKN5folly7dynamicE + 248
10  JitsiMeet                           0x000000010a4e776c ___ZN8facebook5react15RCTNativeModule6invokeEjON5folly7dynamicEi_block_invoke + 88
11  libdispatch.dylib                   0x000000010b943824 _dispatch_call_block_and_release + 24
12  libdispatch.dylib                   0x000000010b944dc8 _dispatch_client_callout + 16
13  libdispatch.dylib                   0x000000010b94ce6c _dispatch_lane_serial_drain + 720
14  libdispatch.dylib                   0x000000010b94db60 _dispatch_lane_invoke + 460
15  libdispatch.dylib                   0x000000010b957bfc _dispatch_workloop_worker_thread + 1220
16  libsystem_pthread.dylib             0x00000001fe6490dc _pthread_wqthread + 312
17  libsystem_pthread.dylib             0x00000001fe64bcec start_wqthread + 4

invite post to use application/JSON

changed to .ajax from .post to allow content type to be set

A reducer must always return a state or null to ignore it. When the color scheme
is undefined we should return the previous state.

Both Chrome and Firefox can work without extensions now, so it no longer makes
sense to default to disabling it in Chrome. Moreover, rely on the fact that
undefined is falsey so no actual config needs to be provided.