Giter Site home page Giter Site logo

ig-lazy-module-loader's Introduction

Ig-Lazy-Module-Loader

Build Status Maven Central License

This library helps with loading modules (features) in Android apps on demand, whenever needed. Before this library can be used a module needs to be compiled to a separate jar/dex or apk file.

Right now, the library supports java libraries and android libraries which don't rely on android resources. Support for lazy loading resources may be added later.

In the demo app there's two examples of lazy loading:

  • java library - compiled first to jar and then to dex file
  • android library - compiled to apk file (in this example I use Android service)

Benefits

Lazy loading of a feature vs having it in the main main executable file brings these benefits:

  • feature is loaded in memory only when really needed. It offloads code from the main executable file which remains smaller which guarantees better cold start time. On Dalvik it offloads methods off the main dex file decreasing performance penalty of multi dex
  • feature code is clustered together in memory as it lives in one file and it provides most optimal execution in terms of memory access
  • less disk space is used if some features remain unused because code is not uncompressed
  • it may improve developer velocity where many developers concurrently work on the same codebase by providing very rigid isolation between features
  • module hotswapping may be implemented allowing for faster development without restarting the app

Tactics around lazy loading

Not every module or a feature is a good candidate to be lazily loaded as lazy loading may incur small latency on the first load.

Features used during cold start would ideally live in the main execubtale file (classes.dex) as this would ensure the most efficient app start. Anything that is outside cold start or, in general, is a less core to the app could be lazily loaded.

There are some tactics around when to trigger lazy loading. Generally speaking a module should be loaded when it is expected to be used in the near future. To make sure the loading latency does not worsen user experience these tactics can be applied:

  • loading a module in the background when user is one click away from the module - it may mean that module remains unused if user doesn't decide to click into that feature or navigates back. But if there's high probability of clicking into that module then it's a well working solution
  • loading a module once user navigates into the module. If loading latency is small (below 50ms) for majority of cases (e.g. p99) then code can block on loading and once it's done navigate into the feature. Otherwise a simple spinner or a progress bar can be displayed so that app does not appear as frozen.
  • some modules are by nature asynchronous and this eases lazy loading as it will be the part of the asynchronous loading. An example of such module from Instagram domain is video player which runs in a secondary process. Instagram initially shows a screenshot of the video while video loads in the background (often times is fetched from the network). Lazy loading would happen in that secondary process and be completely transparent to users.

Also, the first time module is loaded after app install or app upgrade the loading latency will be higher because of dexopt or dex2oat being run (which is compiling and optimizing code for faster execution). This case, although it happens rarely compared to every cold start (e.g. once a week - after app upgade) it usually forces to apply more conservative approach (e.g. loading one click away) especially for modules of significate size (at least couple of hundred kilobytes).

Getting started

The easiest way is to look at the demoapp. It's an Android app that shows a screen with two buttons to lazy load a java library and android library:

Each feature has a Proxy object which lives in the main app. Proxy handles lazy loading of the actual module and delegates calls to it:

class ServiceProxy extends Service {

    private ServiceLike mLazyLoadedService;

    void onCreate() {
        try {
          mLazyLoadedService = LazyModuleLoaderHelper
            .createLoaderWithoutNativeLibrariesSupport(this, new ManifestReader(), mLazyLoadListener)
            .loadServiceModule(ManifestReader.LazyLoadedService, CLASS_NAME);
        } catch (LazyLoadingException e) {
          Log.e(TAG, "Failed to lazy loaded a service", e);
        }
        // delegate calls once lazy loading finished
        mLazyLoadedService.onCreate();
      }
      
      public IBinder onBind(Intent intent) {
        // keep delegating calls
        return mLazyLoadedService.onBind(intent);
      }
      // ...
  }

In the main app an implementation for ModuleManifestReader interface must be added which provides basic metadata about lazy loaded module

class ManifestReader implements ModuleManifestReader {

  public final static String LazyLoadedService = "LazyLoadedService";

  // Those file need to be placed in the assets/ folder
  private final static String ServiceFileName = "lazyservice.apk";

  // Hash of the module may be useful when versioning of modules needs to be added
  private final static String ModuleHash = "null";

  @Override
  public ModuleManifest readModuleManifest(String moduleName) throws IOException {
    if (moduleName.equals(LazyLoadedService)) {
      return new ModuleManifest.Builder(moduleName).setModuleHash(ModuleHash).setDexFileName
          (ServiceFileName).build();
    } else  {
      // ...
    }
  }
}

Each compiled lazy loaded module must be placed in the assets folder. It could either be a dex file (for a java library) or apk (for android library). Right now there's no support for aar files. In the demoapp lazylibrary.dex is compiled file of lazyloadedlibrary and lazyservice.apk is compiled file of lazyloadedservice - both included in this repo:

$ ls LazyLoader/demoapp/src/main/assets/*
src/main/assets//LazyLoadedModule/lazylibrary.dex
src/main/assets//LazyLoadedService/lazyservice.apk

For java library you need to compile the library to a jar file and and then to dex file using dx tool. You also need to add a provided dependency in the gradle build file:

dx --dex --output=lazyloadedlibrary.dex lazyloadedlibrary.jar
dependencies {
    provided project(':lazyloadedlibrary')
}

Compile a AAR

./gradlew clean assembleRelease

Outputs can be found in ig-lazy-module-loader/build/outputs/

Run the Tests

./gradlew clean test

Gradle

Add this to your build.gradle file:

allprojects {
  repositories {
    maven { url 'https://jitpack.io' }
  }
}

dependencies {
    compile 'com.github.instagram:ig-lazy-module-loader:master-SNAPSHOT'
}

Other Instagram Android Projects

Instagram Engineering Blog

License

Copyright (c) 2017-present, Facebook, Inc.
All rights reserved.

This source code is licensed under the BSD-style license found in the
LICENSE file in the root directory of this source tree. An additional grant
of patent rights can be found in the PATENTS file in the same directory.

ig-lazy-module-loader's People

Contributors

juliankrz avatar

Watchers

 avatar  avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.