A simple light-weight analytics wrapper for Swift
Analytical is a simple lightweight analytics wrapper for iOS Swift projects. Inspired by ARAnalytics - a powerful Objective-C analytics library by Orta Therox. Analytical does not support all advanced functionalities of specific providers, but it allows direct access each analytics instance for additional features, after it is set up.
Currently available analytics providers:
A special set of providers:
Analytical is used in production in all applications by Blub Blub.
Analytical is available only through CocoaPods. To install it, simply add the following line to your Podfile:
!use_frameworks
pod 'Analytical'
For performance reasons, it is not recommended to have more than two providers installed, you should specify which you wish to use.
!use_frameworks
pod 'Analytical/Facebook'
pod 'Analytical/Flurry'
pod 'Analytical/Log'
pod 'Analytical/Mixpanel'
pod 'Analytical/Segment'
The Log
subspec will install a simple logger to output tracking calls to Xcode Console using os_log
API.
Analytical includes both Google Analytics and Firebase providers, but due to it’s incompatibility between static libraries and Swift dynamic frameworks, it must be installed manually. To do this there are 5 steps required (sample below is for the GA, but works the same with Firebase):
pod "Google/Analytics"
and pod "Analytical/Core"
to your target’s podfile.pod install
#import <Google/Analytics.h>
to your Application Bridging Header.GoogleProvider.swift
to your project.GoogleProvider
class with Google Tracking ID.Google Analytics takes 4 parameters for each event: Category, Action, Label and Value. Category will be set to “default”, when called from Analytical and Action will be the event name passed to the call. Label and Value are optional.
When Google will update their library to support dynamic frameworks, these steps will no longer be required.
If you do not use CocoaPods, you can manually install Analytical. The required files are in Analytical/Classes/Core
directory, so you may directly drag & drop them into your project: Analytical.swift
, Analytics.swift
and Provider.swift
.
In addition to the core files, you will also require at least one concrete provider, from Analytical/Classes/Provider
directory.
Please ensure that all additional dependencies for specific providers are linked correctly to your target.
To separate analytics code, a new separate Swift file is recommended:
// Define Providers and create a global variable to be accessible from everywhere
let analytics = Analytics() <<~ GoogleProvider(trackingId: "<GA_TRACKING_ID>") <<~ MixpanelProvider(token: "<MIXPANEL_TOKEN>") <<~ FacebookProvider()
// Simple Enum for Events
public enum Track {
public enum Event : String {
case secondScreenTap = "SecondScreenTap"
case closeTap = "CloseTap"
}
public enum Screen : String {
case first = "first"
case second = "second"
}
}
//
// Add simple wrapper to use newly defined enums with Analytical
//
extension Analytical {
func track(event: Track.Event, properties: Properties? = nil) {
self.event(name: event.rawValue, properties: properties)
}
func track(screen: Track.Screen, properties: Properties? = nil) {
self.screen(name: screen.rawValue, properties: properties)
}
func time(event: Track.Event, properties: Properties? = nil) {
self.time(name: event.rawValue, properties: properties)
}
func finish(event: Track.Event, properties: Properties? = nil) {
self.finish(name: event.rawValue, properties: properties)
}
}
Some analytics providers require to be setup when application finishes launching. Make sure to add call to setup
to your application:didFinishLaunchingWithOptions
method:
analytics.setup(with: application, launchOptions: launchOptions)
Some analytics providers require to log application activation (Facebook for example), so you should add the code below to your applicationDidBecomeActive:
method.
analytics.activate()
In both cases ensure that analytics providers are initialized before calling this method.
// Calls using above wrapper
analytics.track(event: .secondScreenTap)
analytics.track(screen: .first)
//
// Original call with String
//
analytics.screen(name: Track.Screen.first.rawValue)
See Example project for more examples.
analytics.track(event: .closeTap, ["property" : 12.00])
analytics.track(screen: .first, ["property" : 12.00])
If your application has identified user, you should call identify
method. If your user is anonymous, you may use analytics.deviceId
property to retrieve UUID. After first retrieval, it is stored to UserDefaults
and used in all next calls. If your application is linked to iAd.framework
, it will return ASIdentifierManager.sharedManager().advertisingIdentifier
as a String
. Analytics does not depend on iAd directly and will use runtime introspection to detect it. It does respect Limit Ad Tracking setting in user’s privacy settings.
analytics.identify(analytics.deviceId)
Where is Carthage support?
Analytical does not support Carthage for providers, as it depends on several external frameworks. You can use the core Analytical as a framework and add specific providers manually, as described under Manual installation.
Dal Rupnik
Analytical is available under the MIT license. See LICENSE file for more information.