This open-source library allows you to build iOS apps compatible with Matrix (http://www.matrix.org), an open standard for interoperable Instant Messaging and VoIP.
This SDK implements an interface to communicate with the Matrix Client/Server API which is defined at http://matrix.org/docs/api/client-server/.
The SDK uses CocoaPods (http://cocoapods.org/) as library dependency manager. In order to set this up:
sudo gem install cocoapods pod setup
The best way to add the last release of the Matrix SDK to your application project is to add the MatrixSDK dependency to your Podfile:
# Obj-C pod 'MatrixSDK' # Swift pod 'SwiftMatrixSDK'
If you want to use the develop version of the SDK, use instead:
# Obj-C pod 'MatrixSDK', :git => 'https://github.com/matrix-org/matrix-ios-sdk.git', :branch => 'develop'
# Swift pod 'SwiftMatrixSDK', :git => 'https://github.com/matrix-org/matrix-ios-sdk.git', :branch => 'develop'
If you want to enable VoIP using the http://webrtc.org VoIP stack, add the following pod to you app Podfile:
pod 'MatrixSDK/JingleCallStack'
As a quick overview, there are the classes to know to use the SDK.
MXRestClient : | Exposes the Matrix Client-Server API as specified by the Matrix standard to make requests to a homeserver. |
---|
These classes are higher level tools to handle responses from a homeserver. They contain logic to maintain consistent chat room data.
MXSession : | This class handles all data arriving from the homeserver. It uses a MXRestClient instance to fetch data from the homeserver, forwarding it to MXRoom, MXRoomState, MXRoomMember and MXUser objects. |
---|---|
MXRoom : | This class provides methods to get room data and to interact with the room (join, leave...). |
MXRoomState : | This is the state of room at a certain point in time: its name, topic, visibility (public/private), members, etc. |
MXRoomMember : | Represents a member of a room. |
MXUser : | This is a user known by the current user, outside of the context of a room. MXSession exposes and maintains the list of MXUsers. It provides the user id, displayname and the current presence state |
The sample app (https://github.com/matrix-org/matrix-ios-console) demonstrates how to build a chat app on top of Matrix. You can refer to it, play with it, hack it to understand the full integration of the Matrix SDK. This section comes back to the basics with sample codes for basic use cases.
One file to import:
Obj-C:
#import <MatrixSDK/MatrixSDK.h>
Swift:
import SwiftMatrixSDK
This API does not require the user to be authenticated. So, MXRestClient instantiated with initWithHomeServer does the job:
Obj-C:
MXRestClient *mxRestClient = [[MXRestClient alloc] initWithHomeServer:@"http://matrix.org"]; [mxRestClient publicRooms:^(NSArray *rooms) { // rooms is an array of MXPublicRoom objects containing information like room id NSLog(@"The public rooms are: %@", rooms); } failure:^(MXError *error) { }];
Swift:
let homeServerUrl = URL(string: "http://matrix.org")! let mxRestClient = MXRestClient(homeServer: homeServerUrl, unrecognizedCertificateHandler: nil) mxRestClient.publicRooms { response in switch response { case .success(let rooms): // rooms is an array of MXPublicRoom objects containing information like room id print("The public rooms are: \(rooms)") case .failure: break } }
Here the user needs to be authenticated. We will use [MXRestClient initWithCredentials]. You'll normally create and initialise these two objects once the user has logged in, then keep them throughout the app's lifetime or until the user logs out:
Obj-C:
MXCredentials *credentials = [[MXCredentials alloc] initWithHomeServer:@"http://matrix.org" userId:@"@your_user_id:matrix.org" accessToken:@"your_access_token"]; // Create a matrix client MXRestClient *mxRestClient = [[MXRestClient alloc] initWithCredentials:credentials]; // Create a matrix session MXSession *mxSession = [[MXSession alloc] initWithMatrixRestClient:mxRestClient]; // Launch mxSession: it will first make an initial sync with the homeserver // Then it will listen to new coming events and update its data [mxSession start:^{ // mxSession is ready to be used // Now we can get all rooms with: mxSession.rooms; } failure:^(NSError *error) { }];
Swift:
let credentials = MXCredentials(homeServer: "http://matrix.org", userId: "@your_user_id:matrix.org", accessToken: "your_access_token") // Create a matrix client let mxRestClient = MXRestClient(credentials: credentials, unrecognizedCertificateHandler: nil) // Create a matrix session let mxSession = MXSession(matrixRestClient: mxRestClient) // Launch mxSession: it will first make an initial sync with the homeserver mxSession.start { response in guard response.isSuccess else { return } // mxSession is ready to be used // now wer can get all rooms with: mxSession.rooms }
We use the same code as above but we add a MXFileStore that will be in charge of storing user's data on the file system. This will avoid to do a full sync with the homeserver each time the app is resumed. The app will be able to resume quickly. Plus, it will be able to run in offline mode while syncing with the homeserver:
Obj-C:
MXCredentials *credentials = [[MXCredentials alloc] initWithHomeServer:@"http://matrix.org" userId:@"@your_user_id:matrix.org" accessToken:@"your_access_token"]; // Create a matrix client MXRestClient *mxRestClient = [[MXRestClient alloc] initWithCredentials:credentials]; // Create a matrix session MXSession *mxSession = [[MXSession alloc] initWithMatrixRestClient:mxRestClient]; // Make the matrix session open the file store // This will preload user's messages and other data MXFileStore *store = [[MXFileStore alloc] init]; [mxSession setStore:store success:^{ // Launch mxSession: it will sync with the homeserver from the last stored data // Then it will listen to new coming events and update its data [mxSession start:^{ // mxSession is ready to be used // Now we can get all rooms with: mxSession.rooms; } failure:^(NSError *error) { }]; } failure:^(NSError *error) { }];
Swift:
let credentials = MXCredentials(homeServer: "http://matrix.org", userId: "@your_user_id:matrix.org", accessToken: "your_access_token") // Create a matrix client let mxRestClient = MXRestClient(credentials: credentials, unrecognizedCertificateHandler: nil) // Create a matrix session let mxSession = MXSession(matrixRestClient: mxRestClient) // Make the matrix session open the file store // This will preload user's messages and other data let store = MXFileStore() mxSession.setStore(store) { response in guard response.isSuccess else { return } // Launch mxSession: it will sync with the homeserver from the last stored data // Then it will listen to new coming events and update its data mxSession.start { response in guard response.isSuccess else { return } // mxSession is ready to be used // now we can get all rooms with: mxSession.rooms() } }
We reuse the mxSession instance created before:
Obj-C:
// Retrieve the room from its room id MXRoom *room = [mxSession room:@"!room_id:matrix.org"]; // Add a listener on events related to this room [room.liveTimeline listenToEvents:^(MXEvent *event, MXEventDirection direction, MXRoomState *roomState) { if (direction == MXTimelineDirectionForwards) { // Live/New events come here } else if (direction == MXTimelineDirectionBackwards) { // Events that occurred in the past will come here when requesting pagination. // roomState contains the state of the room just before this event occurred. } }];
Swift:
// Retrieve the room from its room id let room = mxSession.room(withRoomId: "!room_id:matrix.org") // Add a listener on events related to this room _ = room?.liveTimeline.listenToEvents { (event, direction, roomState) in switch direction { case .forwards: // Live/New events come here break case .backwards: // Events that occurred in the past will come here when requesting pagination. // roomState contains the state of the room just before this event occurred. break } }
Let's load a bit of room history using paginateBackMessages:
Obj-C:
// Reset the pagination start point to now [room.liveTimeline resetPagination]; [room.liveTimeline paginate:10 direction:MXTimelineDirectionBackwards onlyFromStore:NO complete:^{ // At this point, the SDK has finished to enumerate the events to the attached listeners } failure:^(NSError *error) { }];
Swift:
// Reset the pagination start point to now room?.liveTimeline.resetPagination() room?.liveTimeline.paginate(10, direction: .backwards, onlyFromStore: false) { _ in // At this point, the SDK has finished to enumerate the events to the attached listeners }
This action does not require any business logic from MXSession: We can use MXRestClient directly:
Obj-C:
[mxRestClient sendTextMessageToRoom:@"the_room_id" text:@"Hello world!" success:^(NSString *event_id) { // event_id is for reference // If you have registered events listener like in the previous use case, you will get // a notification for this event coming down from the homeserver events stream and // now handled by MXSession. } failure:^(NSError *error) { }];
Swift:
client.sendTextMessage(toRoom: "the_room_id", text: "Hello World!") { (response) in if case .success(let eventId) = response { // eventId is for reference // If you have registered events listener like in the previous use case, you will get // a notification for this event coming down from the homeserver events stream and // now handled by MXSession. } }
In Matrix, a homeserver can send notifications out to a user when events arrive for them. However in APNS, only you, the app developer, can send APNS notifications because doing so requires your APNS private key. Matrix therefore requires a seperate server decoupled from the homeserver to send Push Notifications, as you cannot trust arbitrary homeservers with your application's APNS private key. This is called the 'Push Gateway'. More about how notifications work in Matrix can be found at https://matrix.org/docs/spec/push_gateway/latest.html
In simple terms, for your application to receive push notifications, you will need to set up a push gateway. This is a publicly accessible server specific to your particular iOS app that receives HTTP POST requests from Matrix Home Servers and sends APNS. Matrix provides a reference push gateway, 'sygnal', which can be found at https://github.com/matrix-org/sygnal along with instructions on how to set it up.
You can also write your own Push Gateway. See https://matrix.org/docs/spec/push_gateway/latest.html for the specification on the HTTP Push Notification protocol. Your push gateway can listen for notifications on any path (as long as your app knows that path in order to inform the homeserver) but Matrix strongly recommends that the path of this URL be '/_matrix/push/v1/notify'.
In your application, you will first register for APNS in the normal way (assuming iOS 8 or above):
UIUserNotificationSettings *settings = [UIUserNotificationSettings settingsForTypes:(UIRemoteNotificationTypeBadge |UIRemoteNotificationTypeSound |UIRemoteNotificationTypeAlert) categories:nil]; [...] - (void)application:(UIApplication *)application didRegisterUserNotificationSettings:(UIUserNotificationSettings *)notificationSettings { [application registerForRemoteNotifications]; }
When you receive the APNS token for this particular application instance, you then encode this into text and use it as the 'pushkey' to call setPusherWithPushkey in order to tell the homeserver to send pushes to this device via your push gateway's URL. Matrix recommends base 64 encoding for APNS tokens (as this is what sygnal uses):
- (void)application:(UIApplication*)app didRegisterForRemoteNotificationsWithDeviceToken:(NSData*)deviceToken { NSString *b64Token = [self.deviceToken base64EncodedStringWithOptions:0]; NSDictionary *pushData = @{ @"url": @"https://example.com/_matrix/push/v1/notify" // your push gateway URL }; NSString *deviceLang = [NSLocale preferredLanguages][0]; NSString *profileTag = makeProfileTag(); // more about this later MXRestClient *restCli = [MatrixSDKHandler sharedHandler].mxRestClient; [restCli setPusherWithPushkey:b64Token kind:@"http" appId:@"com.example.supercoolmatrixapp.prod" appDisplayName:@"My Super Cool Matrix iOS App" deviceDisplayName:[[UIDevice currentDevice] name] profileTag:profileTag lang:deviceLang data:pushData success:^{ // Hooray! } failure:^(NSError *error) { // Some super awesome error handling goes here } ]; }
When you call setPusherWithPushkey, this creates a pusher on the homeserver that your session is logged in to. This will send HTTP notifications to a URL you supply as the 'url' key in the 'data' argument to setPusherWithPushkey.
You can read more about these parameters in the Client / Server specification (http://matrix.org/docs/api/client-server/#!/Push32notifications/post_matrix_client_r0_pushers_set). A little more information about some of these parameters is included below:
- appId
- This has two purposes: firstly to form the namespace in which your pushkeys exist on a homeserver, which means you should use something unique to your application: a reverse-DNS style identifier is strongly recommended. Its second purpose is to identify your application to your Push Gateway, such that your Push Gateway knows which private key and certificate to use when talking to the APNS gateway. You should therefore use different app IDs depending on whether your application is in production or sandbox push mode so that your Push Gateway can send the APNS accordingly. Matrix recommends suffixing your appId with '.dev' or '.prod' accordingly.
- profileTag
- This identifies which set of push rules this device should obey. For more information about push rules, see the Client / Server push specification: http://matrix.org/docs/api/client-server/#!/Push32notifications/post_matrix_client_r0_pushers_set This is an identifier for the set of device-specific push rules that this device will obey. The recommendation is to auto-generate a 16 character alphanumeric string and use this string for the lifetime of the application data. More advanced usage of this will allow for several devices sharing a set of push rules.
The repository contains a Xcode project in order to develop. This project does not build an app but a test suite. See the next section to set the test environment.
Before opening the Matrix SDK Xcode workspace, you need to build it.
The project has some third party library dependencies declared in a pod file. You need to run the CocoaPods command to download them and to set up the Matrix SDK workspace:
$ pod install
Then, open MatrixSDK.xcworkspace
.
The tests in the SDK Xcode project are both unit and integration tests.
Out of the box, the tests use one of the homeservers (located at http://localhost:8080) of the "Demo Federation of Homeservers" (https://github.com/matrix-org/synapse#running-a-demo-federation-of-synapses).
You first need to follow instructions to set up Synapse in development mode at https://github.com/matrix-org/synapse#synapse-development. If you have already installed all dependencies, the steps are:
$ git clone https://github.com/matrix-org/synapse.git $ cd synapse $ virtualenv -p python2.7 env $ source env/bin/activate (env) $ python synapse/python_dependencies.py | xargs pip install (env) $ pip install lxml mock
Then, you need to set Synapse up for running demo scripts (https://github.com/matrix-org/synapse/blob/master/demo/README):
(env) $ python setup.py develop
Installation is done.
Every time you want to launch these test homeservers, type:
$ virtualenv -p python2.7 env $ source env/bin/activate (env) $ demo/start.sh --no-rate-limit
You can now run tests from the Xcode Test navigator tab or select the MatrixSDKTests scheme and click on the "Test" action.
CocoaPods may fail to install on OSX 10.8.x with "i18n requires Ruby version >= 1.9.3.". This is a known problem similar to CocoaPods/CocoaPods#2458 that needs to be raised with the cocoapods team.
### Dynamic Framework: Undefined symbols for architecture
If you are using "MatrixSDK" instead of "SwiftMatrixSDK", you may get a compile- time error that looks like this:
Undefined symbols for architecture x86_64: "_OBJC_CLASS_$_GAIDictionaryBuilder", referenced from: objc-class-ref in MXGoogleAnalytics.o "_OBJC_CLASS_$_GAI", referenced from: objc-class-ref in MXGoogleAnalytics.o ld: symbol(s) not found for architecture x86_64 clang: error: linker command failed with exit code 1 (use -v to see invocation)
This happens when both of the following are true:
1. The MatrixSDK was compiled as a framework (i.e. the use_frameworks! setting is enabled in your podfile) 2. Your project also uses the Google Analytics pod.
The root cause is that the MXGoogleAnalytics class recognizes that the Google Analytics pod was included in your project, and attempts to include its headers. This type of behavior is allowed in a Static Library, but it is not allowed in a Dynamic Framework.
The easiest workaround is to switch to the "SwiftMatrixSDK" pod, even if you don't use Swift (for an Obj-C project, the pods are virtually the same). The "SwiftMatrixSDK" excludes the MXGoogleAnalytics class. If you want to collect analytics data about initialization times (and so forth), you can implement your own MXAnalyticsDelegate and set an instance to MXSDKOptions.sharedInstance. See MXAnalyticsDelegate.h and MXGoogleAnalytics.h/m for more information.
The SDK currently manages only login-password type registration. This type of registration is not accepted by the homeserver hosted at matrix.org. It has been disabled for security and spamming reasons. So, for now, you will be not be able to register a new account with the SDK on such homeserver. But you can login an existing user.
If you run your own homeserver, the default launch parameters enables the login-password type registration and you will be able to register a new user to it.
Copyright (c) 2014-2017 OpenMarket Ltd Copyright (c) 2017 Vector Creations Ltd Copyright (c) 2017-2018 New Vector Ltd
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this work except in compliance with the License. You may obtain a copy of the License in the LICENSE file, or at:
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.