Tracklytics 一个安卓埋点库

Tracklytics 一个安卓埋点库

声明：该库是由 https://github.com/orhanobut/tracklytics 作者的埋点库修改而来
推荐先看下原作者的文档先，他写的更详细，我的仅仅是修复了问题及导入方式不同而已，使用上依然是一样的。

修复了如下问题：
1、解耦了Aspectjx AOP库
2、修复无法跨库使用的问题

1、导入

在你的项目主build.gradle中加入aspectjx依赖

buildscript {
    dependencies {
        classpath 'com.hujiang.aspectjx:gradle-android-plugin-aspectjx:2.0.10'
    }
}

在所需要用到埋点的模块module/build.gradle中引用aspectjx的插件

apply plugin: 'com.hujiang.android-aspectjx'

以及引用依赖

dependencies {
    implementation 'com.ubitar.tracklytics:tracklytics:1.0.0'
}

2、使用

注册埋点监听器，每当出发埋点时将会调用

Tracklytics.init(new EventSubscriber() {
  @Override public void onEventTracked(Event event) {
    // Send your events to Mixpanel, Fabric etc
  }
});

添加埋点事件

@TrackEvent("event_name") 
public void foo() {
}

3、常见问题

编译时出现 zip file is empty
参考https://github.com/HujiangTechnology/gradle_plugin_android_aspectjx/issues/287 这个是Aspectjx的版本问题，多去issues里搜搜解决办法

4、更多使用功能（从这里我就直接搬运了，太多了）

@TrackEvent

Scope: Method

Track an event and notify the subscriber(EventSubscriber) on each method invocation.

@TrackEvent("event_name") 
public void foo() {
}

@Attribute

Scope: Method, Method parameters

There are multiple ways to add an attribute to the corresponding event. Assigns the values in runtime dynamically.

By using method parameters: Parameter value will be used as attribute value.

@TrackEvent("event_name") 
public void foo(@Attribute("attribute_key") String name) {
  // something
}

By using the return value of the method as attribute value.

@TrackEvent("event_name") 
@Attribute("attribute_key")
public String foo() {
  // something
  return "attribute_value";
}

Set a default value when the expected value is null.

@TrackEvent("event_name") 
public void foo(@Attribute(value="attribute_key", defaultValue="defaultValue") String name) {
  // something
}

@FixedAttribute

Scope: Method, Class

If the attribute values are constant, use FixedAttribute.

On method: Only this event will have this fixed attribute.

@TrackEvent("eventName")
@FixedAttribute(key="Login", value="Success")
public void foo(){
}

On class: These attributes will be added to each event that is triggered within this class. For example: Following foo() method will also have screen_name attribute.

@FixedAttribute(key="screen_name", value="Login")
public class LoginPresenter{

  @TrackEvent("login")
  public void onLoggedIn(){
  }
}

@FixedAttributes

Scope: Method, Class

Prior to Java 8, repeated annotations are not available. Sometimes you may need more fixed attributes. Use this annotation to add multiple attributes

@TrackEvent("event_name")
@FixedAttributes({
  @FixedAttribute(key="name", value="Something"),
  @FixedAttribute(key="last_name", value="Something")
})
public void foo(){
}

Filters

Use filters to differentiate some events. You may only want to send specific events to a specific analytics tool. ie: Send login event to Fabric.

@TrackEvent(value="event_name",filters=100)
public void foo() {
}

Tags

You can use tags to send more information about the tracked event. For example: Adjust requires token for their events.

@TrackEvent(value="event",filters=100, tags="abc123")
public void trackNoValues() {
}

Super Attributes

Some attributes might be used for every event within the app such as device id. Tracklytics call them as super attributes. These attributes will be automatically added to each event. You only need to set them once and Tracklytics will do the rest.

Access super attributes via Event class

Tracklytics.init(new EventSubscriber() {
  @Override public void onEvent(Event event) {
    // event.superAttributes
  }
});

Set any attribute as super

@Attribute(value="key", isSuper=true)

Set any fixed attribute as super

@FixedAttribute(key="key", value="value",  isSuper=true)

Add super attribute directly

tracklytics.addSuperAttribute("key","value");

Remove super attribute

tracklytics.removeSuperAttribute("key");

@TrackableAttribute / @Trackable

You can make the class trackable in order to provide preset values. Imagine your domain models, they contain a lot of information and you can re-use them as a source for tracking. For the following use case: When event is triggered, attributes from Foo.getTrackableAttributes() will be added to this event.

class Foo implements Trackable {
  String name;

  @Override public Map<String, String> getTrackableAttributes() {
    Map<String,String> values = new HashMap<>();
    values.put("name", name);
    return values;
  }
}

@TrackEvent("event")
void something(@TrackableAttribute FooTrackable foo){
}

@TransformAttribute / @TransformAttributeMap

Sometimes values are not in the correct form, such as position or index is in Integer type. This might not be clear if you track them as raw. You may want to send a more understandable value to the analytic tools. Use TransformAttribute to solve this issue.

For example: In the following example, index is represented by integer and you want to have a String value which represent exact value such as menu item.

class Foo {
  @TrackEvent("event")
  @TransformAttributeMap(
    keys = {0, 1},
    values = {"value0", "value1"}
  )
  public void foo(@TransformAttribute("key") int index) {
  }
}

// foo(0) : event -> [{"key","value0}]
// foo(1) : event -> [{"key","value1}]

Log

Tracklytics provides a stream for logs which sends formatted event messages.

Tracker tracklytics = Tracker.init(...);
tracklytics.setEventLogListener(new EventLogListener() {
  @Override public void log(String message) {
    // Set your logger here. ie: Logger or Timber
    Log.d("Tracker", message);
  }
});

Log output sample

eventName:{key=value}, super attrs: {key=value}, tags={100,200}

More API options

You can also track event directly without annotations.

tracklytics.trackEvent(String eventName)

tracklytics.trackEvent(String eventName, Map attributes)