Learn about Pokémon Go.

发表于2017-08-02
评论0 1.6k浏览

Recently  Pokémon Go  is too red, plus their own technical control, see this article  " Unbundling Pokémon Go "    : how to use reverse engineering to get  app  source code, and analyze its operating mechanism, in this translation to share with everyone


This translation has been  approved  by  Adrien Couque  , as follows  :  

I do not know where to come from now,  Pokémon Go  swept the world in a week, and we found something interesting from it

Although but at The  App  IS Rate this page Currently only the Available in Three the Countries  (  US, Australia and New Zealand  )  ,  IT Still Makes  Twitter and Facebook Go Pale     .  It  Defeated  Candy Crush to Become at The MOST successful Mobile Game in at The United States, not only proved  to the bring Benefits  To  developers  , local businesses also pay attention to  Pokémon Go  will bring those  tourists  , Nintendo's market value and then increased by  90%  . 

The game in such a short time to become a household name, inspired we want to see its internal structure.  This article takes  Pokémon Go App  as an example to how to get the  Android App  program code from  reverse engineering  and analyze its network connection request for More information


Prepare  APK

Before you can do reverse engineering, you must have an  APK  file, and it  is difficult to  get  Pokémon Go  's  APK  file.  Please note that installing  APKs  with  unknown sources  will be a high security risk, but  Google Play  will  do some analysis to the reduce at The Risk of  App  , SO IT IS Best for people to  download and install  App  through  Google Play  .  But for Reverse Engineering, at The MOST like THESE Malicious  APK  , Because IT IS interesting.  Here, WE are  ANALYZING at The  Release of  Pokémon Go  Version  0.29.0  for  7/7  .

First talk about, do reverse engineering, we still do not see some things:

·           The Any now Build code -related Things  

·           The Any with the Continuous Integration and Testing (the Continuous Integration) Things  

·           Other Special versions ( EG Debug Version ) : During Development, you May have have some Special Features But by Will not BE Placed in at The Final Product, Because WE want to the Analyze at The Pokémon Go Version of at The Official Release, SO at The Special function Should by Will not Be put inside this        

The           code for the backend service.  Many people may want to know how the algorithm determine the location where the Pokemon appears.  But this algorithm is on the back end, we can only know how to transfer data with the back, can not know the logic Of the internal algorithm


APK  content

Let's take a look at  the internal structure of  APK  .  In fact,  APK  is just a  zip  archive, which contains:

Here's a description  of the function of  each file  (  green  )  and folder  (  red  )  :

·          Manifest is the Android Manifest , it's like App 's identity card, which provides information on the name, icon, version, privileges, restrictions and other hardware components. This is also required when the system installs or upgrades the App . 

After the          program is compiled on classes.dex , you can have more than one classes.dex .

The          lib file is loaded with a library

The          res and assets archive is a static resource file

·           Resources.arsc  IS Android A Special File, at The Compiler R.java Generated, Which IS and static Used to Link Resources.    

·           META-INF  folders filled with the intermediary data (the Metadata) , but here we do not need.   

Above is what you  would see  when you decompress  APK  .

We started to look at the first file:  classes.dex  .


Decompile the program code

De x   is the  Dalvik Executable   abbreviations  (  the Dalvik  is  Android  system legacy virtual machine, now called new  the ART  , the full name is  Android Runtime  , but still use the file extension  dex  )  .  This is the  Android  system-specific file format, and it Is not easy to read its contents.  There are two ways to do this: the first use of  smali  disassembler will  dex  file content into easy to read  bytecode  , the second use  dex2jar  content into traditional  Java  files.        

We intend to use the second method to convert  dex  into  jar  files  (jar  is a compressed file that contains all the   .class  files  )  .  Next we need to decompile the tool and then convert the   . Script  file into  Java  program code.  There are a Lot of ready-made anti-compiler tools, have their own  Jayx  , you can use your usual, and even can find the  online version of the anti-compiler  . 

We now have most of the easy-to-read Java program code, limited by the limitations of the decompiler program, and still some part of the program code can not be seen. In fact, there is a reverse compiler Procyon , which may have better output results. 

It's important that the code we get is not the source code written by the original developer, just like using  Google  Translate to turn English into French and then turn back to english, you will get another string of new English.  The reason is that when you Want to turn into French, according to the contents of the English word or phrase will determine the best corresponding word or sentence, once again turned back to English, according to the contents of the French will do a decision of the best best word or sentence Operation, the process of the back and forth are independent, the result will be different.  This works like the reverse code of the program code: we decompile the program code, its operation will be the same as the original code,But the program code content will not be exactly the same as the original code, the difference may have a function name, variable name And annotations.

Fortunately, we can clearly understand the  app  used in the library:

·           Android support libraries   : support-v4, appcompat and support-annotations

·           Various Parts of AT at The   Play Services

·           Jackson   (JSON parser): core, annotations and databind

·           Gson   (JSON parser)

·           Otto   (event bus)

·           Dagger   (dependency injection)

·           RxJava   /   RxAndroid   (reactive Programming)

·           Apache   Commons IO   (utilities for I / O)

·           AdMob, now declared as   firebase-ads   (ads, analytics)

·           Upsight   (analytics)

·          Crittercism , now known as Apteligent (monitoring and crash reporting)

·          Unity classes.jar (interact between the Android framework and Unity)

·          Lunar Mobile Console  (Unity logger for Android)

·          Voxelbusters's  Cross Platform Native Plugins  (mainly used for sharing from Unity)

·          Google SDK VR

If you are Android developers, you can feel strange: why there are two JSON parser ? One does the active programming ( annotation: author Ray Shih 's opinion on the relevant programming ) , a do event bus This is actually transitive dependencies : the library will be dependent on the operation, but write the program sometimes only call to several of the library, you can get here to understand how we analyze transitive dependencies . 

Clean up some of the library without call, get a more concise list:

·          Gson

·          Crittercism

·          Upsight

·          Admob / firebase-ads

·          Google VR SDK, Unity and associated

Another kind of dependency is from outside to inside, layers of wrapping up, like Upsight inside a large number of libraries, lists the list and the number of functions: RxAndroid (4k), Dagger (~ 200), Commons IO (1k), Jackson (10k), Otto (~ 50), various services (12k), the development of their own functions (3k) .

--- com.upsight.android:all:4.1.3  | --- io.reactivex: rxandroid: 1.0.1  | | --- io.reactivex: rxjava: 1.0.13  | --- com.upsight.android :  Analytics : 4.1.3 | | --- io.reactivex: rxandroid: 1.0.1 (*)  | | --- com.google.dagger: dagger: 2.0.2  | | Javax.inject: javax.inject: 1  | | --- com.upsight.android:core:4.1.3  | | Io.reactivex: rxandroid: 1.0.1 (*)  | | | Com.google.dagger: dagger: 2.0.2 (*)  | Commons-io: commons-io: 2.4  | | | com.fasterxml.jackson.core: jackson-databind: 2.6.3  | | | com.fasterxml.jackson.core: jackson-annotations: 2.6.0  | | | --- com.fasterxml.jackson.core: jackson-core: 2.6.3  | | | com.squareup: otto: 1.3.8  | | --- commons-io: commons -io:2.4  | | --- com.fasterxml.jackson.core: jackson-databind: 2.6.3 (*)  | | \ --- com.squareup: otto: 1.3.8  | --- com.google.dagger: dagger : 2.0.2 (*)  | --- com.upsight.android:google-advertising-id:4.1.3  | | --- io.reactivex: rxandroid: 1.0.1 (*)  | | --- com. Upsight.android:analytics: 4.1.1 (*)  | | --- com.google.dagger: dagger: 2.0.2 (*)  | | --- com.android.support:support-v4:23.2.1 ( *)  | | --- com.google.android.gms: play-services-ads: 8.4.0 -> 9.2.0 (*)  | | --- com.upsight.android:core:4.1.3 (* )  | | --- com.upsight.android:marketing:4.1.3  | | Io.reactivex: rxandroid: 1.0.1 (*)  | | | com.upsight.android:analytics:4.1.3 ( *)  | | | Com.google.dagger: dagger: 2.0.2 (*)  | | | Com.upsight.android:core:4.1.3 (*)  | | | Commons-IO: Commons-IO: 2.4  | | | --- com.fasterxml.jackson.core: jackson-the DataBind : 2.6.3 (*)  | | | com.squareup: otto: 1.3.8  | | --- commons-io: commons-io: 2.4  | | --- com.fasterxml.jackson.core: jackson -databind: 2.6.3 (*)  | | \ --- com.squareup: otto: 1.3.8  | --- com.upsight.android:google-push-services:4.1.3  | | --- io. Reactivex: rxandroid: 1.0.1 (*)  | | --- com.upsight.android:analytics: 4.1.1 (*)  | | --- com.google.dagger: dagger: 2.0.2 (*)  | --- com.android.support:support-v4:23.2.1 (*)  | | --- com.google.android.gms: play-services-gcm: 8.4.0 -> 9.2.0 (*)  | | --- com.upsight.android:core:4.1.3 (*)  | | --- com.upsight.android:marketing:4.1.3 (*)  | | --- commons-io: commons-io: 2.4  | | --- com.fasterxml.jackson.core: Jackson-databind: 2.6.3 (*)  | | \ --- com.squareup: otto: 1.3.8  | --- com.upsight.android:managed-variables:4.1.3  | | --- io.reactivex : Rxandroid: 1.0.1 (*)  | | --- com.upsight.android:analytics: 4.1.1 (*)  | | --- com.google.dagger: dagger: 2.0.2 (*)  | - com.upsight.android:core:4.1.3 (*)  | | --- commons-io: commons-io: 2.4  | | --- com.fasterxml.jackson.core: jackson-databind: 2.6.3 (*)  | | \ --- com.squareup: otto: 1.3.8  | --- com.upsight.android:marketing:4.1.3 (*)  | --- com.upsight.android:core:4.1. 3 (*)  | Commons-io: commons-io: 2.4  | --- com.fasterxml.jackson.core: jackson-databind: 2.6.3 (*)  | \ --- com.squareup: otto: 1.3.8

This means that you have thousands of functions to analyze.

Although the library is a lot, but remove the analysis tools, monitoring tools, crash return and advertising, the most important left Pokémon Go game engine Unity . This is why you open the app will have a Niantic logo, in order to allow users to wait for a moment to start the Unity engine, and then a progress bar, the display engine to read the status of static files. All of your interactive operations are in the implementation of Unity environment, so will not see any Android native interface.

Another Note IS:  the VR the SDK  .  The In  The Stage of  Pokémon Go Beta  , someone Used The Same Way WE found  Cardboard / the VR  and OTHER words in The Program code, in The Official Version of The  App  use Statement Also Mentioned  Cardboard  .  But from My Analysis , The future does not have  the corresponding function of  VR  or  Cardboard  .  From our professional point of view,  VR SDK  this library is used to connect the  Android framework  and  Unity  , but if you really want to  integrate  with the  Cardboard  ,You must let the  Android framework  and  Unity  can communicate with each other, it must refer to a large number of open source Can do it.  But we do not see from the current code.   

Here we spend a lot of time cleaning up the program, but there is not a real implementation of the project, but also need  resources  and  assets  , let us continue to look down.


Static resource file

To get resources and assets is also simpler than the original code. In fact, the assets will be packed into the App , almost all of the assets are used in Unity , so we temporarily regardless of them. Resources are interesting, they include icons layouts and wording The contents of Resources will become easy to read or edit after build . For example, the xml layouts file will be converted to binary format, and the 9-patches file will lose the basis for interpretation. 

The good news is that there is a tool called  apktool  , which can help us to  Manifest  and  resources  files into easy-to-read format content, and produce an executable  Android  project.  At first we didnt use  apktool  will be  classes.dex  into  Smali  file, rather than we want the  Java  program code. 

Now with the decompiled  resources  and  Manifest  , there are also  assets  , plus the code will be cleaned up earlier, we can start to build and implement a complete  Android  project.


Compile and execute

In order to generate  APK  , we have to compile the  Java  program code before the need to create an  Android  project and  build  instructions.  If you remember, because these things are not in  APK  , so we have to own, rely on:  Gradle  .

One of the interesting things is  "the  minimum  demand Android  version of the demand  .  "  App  's minimum demand  on  Google Play  is  Android KitKat (Android 4.4, API level 19)  , but in the analysis of the library, the  maximum demand for  Google VR SDK  is only to  API level 16 (JellyBeans, or Android 4.1)  , we do not Clearly so  Google Play  claims are higher than the actual  API  requirements  3  versions.  This is the beginning to exclude  20%  of  Android  users  (  according to  Google ' S latest numbers  )  , maybe intentional, and maybe a mistake. 

But the most important thing now is that we already have a project that can be executed on the phone. If you want to install this reverse engineering version of the App , it is recommended in your build.gradle and Manifest components / permissions which first replace the application id , to avoid the conflict with the official version to ensure that the official version can be updated at any time.

After the installation is successful, you will find that you are stuck in the login screen. The first sign-in option is Google Sign-In . But when you click on it, it will verify the certificate signed by App , obviously we do not have credentials, so the error message is jumped out: GoogleAuthException: INVALID_AUDIENCE In order to avoid this restriction, we have to spend a lot of effort to have a way, so the easiest way is to directly to the Google Developer Console to apply for a new App , so reverse engineering version of App can have their own credentials, and after login Get token , but still can not do with the back-end data exchange. 

The second login option is through the  Pokémon Trainer Club  application account.  But because too many people apply, the server seems to have been closed, and so it is restored, we will try to see if the reverse engineering version of  App  can log in.  


Analyze the program code

Here we will examine look at the program code.  Since this article is about an overview of reverse engineering, we will focus on  Pokémon Go App  , and  the analysis of  each  app  may not be the same

We saw most of the program code earlier in the  Unity  engine, because  Unity  is cross-platform, so these code can be executed on  iOS  and  Android  .  But some are based on  Android  native features, such as:

·           Sign-in / Registration (Inside at The Package Penalty for AT Penalty for com.nianticlabs.nia.account)

·           In-App purchases (inside com.nianticlabs.nia.iap)

·           Interaction with Location, Network and Sensors (inside com.nianticlabs.nia.location / network / sensors)

·           Communication via Bluetooth with the   Pokémon Go Plus   (inside com.nianticproject.holoholo.sfida)

At first glance the most interesting thing is location / network / sensors program code if you fake your position or speed, the first time to know the location and type of appearance, and then you can catch more Pokémon words ...)

Communicate with Pokémon Go Plus , it should be when your mobile phone on the backpack or pocket when you can tell you near the Pokémon. This part of the code can be combined with the analysis of the network request to allow App only to inform you of the Pokemon you are interested, for example, you have not collected the only.

Slightly look at the code that communicates with Pokémon Go Plus :

Boolean notifyCancelDowser ( ); Boolean notifyError ( ); Boolean notifyFoundDowser ( ); Boolean notifyNoPokeball ( ); Boolean notifyPokeballShakeAndBroken (String STR ); Boolean notifyPokemonCaught ( ); Boolean notifyProximityDowser (String STR ); Boolean notifyReachedPokestop (String STR ); Boolean notifyReadyForThrowPokeball (String Str ); boolean notifyRewardItems (String str );Boolean notifySpawnedLegendaryPokemon (String str ); boolean notifySpawnedPokemon (String str ); boolean notifySpawnedUncaughtPokemon (String str ); boolean notifyStartDowser ( ); 












This is very valuable data!  You can build your own device:


Intercept network connection

Do reverse engineering does not mean that it will take a lot of trouble to dismantle the code, you can from the  App  how to interact with the outside world, this method applied to any software.

App  will be linked with the screen, to do the display or touch the interaction, in addition to: file system, sensors, networks and so on.

Here we are most interested in the network request.  As we resume earlier, the most important game of logic operations are on top of the server,  App  need to exchange data with the server to do you can operate, if we can capture information on These transmissions, we might not have to pass through  App  can and communication.

An In FACT,  Pokémon Go  Used A  Method,  Called  Optimistic Models  the when Dealing with Network Requests  .  Optimistic Models  the allow the Users  to do AN Action  ON at The  App  , do not need to the wait for at The Server's the Response, Directly to at The the Next Action to the Continue to Operate, SO That the player feel good smooth.  If the server later error, it will jump out of the warning.  So you can see when you are in the delivery of Pokémon, and did not show any waiting tips.  App  is currently  in this mechanism has not been Very smooth operation, mainly because the server is loaded, I believe the next few weeks will improve.  

So how do we fetch network requests?  The easiest way is to have  a  proxy in the  middle  of the  App  and the server  .  But if the data is  encrypted  by  HTTPS  , you can only see irre something  metadata  .

There are a way called  Man-in-the-Middle  attack.  This way you use  proxy  to deceive  App  you are  Server  , and then cheat  Server  you are  App  .  When you receive the  App  request, use your  app-side key  first decrypt, And then  server-side key  encryption to  Server to  get a response, and then  decrypt the  server-side key  , and then  app-side key  encryption back to  App  .  You can get complete information, and  App  and  Server  will not know your presence The  

Obviously, if the story is over, all the data on the network will be seen light.  In fact, the use of encryption and decryption  key  is required to be third-party verified, that  Certificate Authorities  .  Your phone or browser will only trust the Verified  key  , or back out of the warning message.  Because the phone is our own, we can put the  first  key  on the phone, to extract the information. 

There are ready-made tools to help us complete the  proxy  settings, like  mitmproxy  and  Charles  .  Charles has  to pay, but the use of interface can guide us to make  settings  .  The following figure is  the Internet request that was intercepted when the  App was Started   

From which you can learn a lot of things, to see the first few requests:

·           Https://android.clients.google.com/c2dm/register3:  Sign up up notifications 

·           Https://stats.unity3d.com/HWStatsUpdate.cgi:  Probably A Unity- Related the Analysis Event  

·           Https://bootstrap.upsight-api.com/config/v1/a9cc12f87adc420baf964f187672ecb4/: Upsight  'S First the Analysis Event

·           Https://appload.ingest.crittercism.com/v0/appload: Crittercism  's first analysis event

·           Https://pgorelease.nianticlabs.com/plfe/rpc: at The  the Details Described below by Will BE

·           Https://play.googleapis.com/log:  Communicate  with Play Services backend  

·           Http://lh4.ggpht.com/LakctgAXpXwe-3PMCWws8rCoVn1_TmyfAiWjWXm6VtsRjRl5v53n1JrWBumWmldzsBFxIUdRLXgsMewLjuyN:  This is an Picasa request is PokéStop pictures    

·           Https://e.crashlytics.com:  Communicate  with Crashlytics , But Seems to IT BE failing  

·           Https://www.google.com/loc/m/api: GPS  location

We can see that  App  frequently  communicates  with  https://pgorelease.nianticlabs.com/plfe/  , and a  226  number is  followed  by the  URL  , and I guess it's done for  load balancing  : that's the first request is assigned to a server Go, then all the same  session  in the same  request will lead the same server.

Finally, the  "rpc"  in the  URL of the  last thing on behalf of the  app  is through the  Remote Procedure Call  to communicate  with the  Server  , so all the requests are sent to the same  URL  , which  is not the same  with the  REST  way The   

Look at the contents of the request, neither  JSON  nor  XML  , and no compression or encryption: so we can clearly see  UUIDs  and  "pm0015"  and other strings, which may use using  protocol buffers   (  or  flat buffers  )  Do the serialized format.  Charles  Will help straighten it out, you can use  protocol buffers  in  the Command Line  , so from:    

5, € â € "ÉßÛS # pgorelease.nianticlabs.com / plfe / 226: [ 
@ nrÝZ † ¡ ϯ½" 'ëXÖÐ _} ?? Î ~ -0' @ ... Ít '> - C ÷ ‰

Sorted into:

1: 53 
2: 6032429073588813826
3: "pgorelease.nianticlabs.com/plfe/226"
7 {
  1: "nr \ 026 \ 335Z \ 206 \ 241 \ 317 \ 257 \ 275 \ 224 \ '\ 353X \ 326 \ 320_} \ 220
  \ 316 ~ \ 227 \ 361 \ 3670 \ '@ \ 205 \ 315t \ 221 \ 233-C \ 367 \ 211 \ r <j8y \ 024
  \ 224 \ 312v \ 342 \ 2269 ~ \ 304 \ 202 / \ 036 \ 247 \ 276 \ 361 \ 266, \ 033s \ 027 \ 006 \ f ^ "
  2: 1468599616357
  3:" $ \ 002 \ 304 \ 337. \ 034 \ 270 \ 361 \ 214D \ 251nz \ 273fM "
}
100 {
}
100 {
} </ j8y \ 024

This is the request that  pgorelease.nianticlabs.com/plfe/rpc  returns, which has a new request endpoint:  pgorelease.nianticlabs.com/plfe/226  , is used for all times requests.

You can also see a lot of  "\ xxx"  , which is  "octal escaping"  .  Use the  decoder  , the content from:

Nr \ 026 \ 335Z \ 206 \ 241 \ 317 \ 257 \ 275 \ 224 \ '\ 353X \ 326 \ 320 \} \ 220 
\ 316 ~ \ 227 \ 361 \ 3670 \' @ \ 205 \ 315t \ 221 \ 233- C \ 367 \ 211 \ r <j8y \ 024
\ 224 \ 312v \ 342 \ 2269 ~ \ 304 \ 202 / \ 036 \ 247 \ 276 \ 361 \ 266, \ 033s \ 027 \ 006 \ f ^ </ j8y \ 024

Becomes

Nr5Z617754 \ '3X60_} 06 ~ 7170 \' @ 55t13-C71 \ r

From the results speculated that this is like a list of objects that appear in the vicinity of Pokémon, each object has its own  UUID  and attributes  (  eg  pm0015  stands for  pokémon 015  :  Beedrill  )  , the other may be coordinates, combat effectiveness and statistics.  We can  prove this  from the request  https://storage.googleapis.com/cloud_assets_pgorelease/bundles/android/pm0126  , if this request can get  pm0126  related  assets  .   

Continue to see the return of other requests.  For example, the bottom of this should be the player's information:

{100 
  . 1 :.. 1.
  2 {
    . 1:. 1,467,925,951,134
    2: "redacted: Player name"
    . 7:. "\ 000 \ 001 \ 003 \ 004 \ A"
    . 8 {.
      . 8 :.. 1.
    }
    . 9:. 250
    10 : 350.
    11 {
    }
    12 is {
    }
    13 {
    }
    14 {
      1: "POKECOIN"
    }
    14 {
      1: "STARDUST"
      2: 500
    }
  }
}

Number The at The  1467925951134  IS  Unix timestamp  , the Refers to at The  07/07/2016 21:12  , the this Should BE at The Player's Registration Time.  An In at The Request and return at The Content, you CAN See Everywhere  timestamp  , some Accuracy to  millisecond  , some to  Nanosecond  .

And then further, we can see a lot of pairs of numbers, such as:  0x40486ddc40000000, 0x4002d99520000000  .  This should be a coordinate, but not encoded in hexadecimal, but  IEEE 754 doubles  .  The value of the hexadecimal value is converted into numbers:

Is the coordinates of our office!  We will be able to get all the coordinates, guess its meaning, are marked on the map:  the position of the user (  yellow  ), points of interests / PokéStops (  red  ) and possible spawn points (  green  )

So far, we will read the network exchange of data, serialized format, but also to distinguish some  id  ,  timestamps  and  GPS  coordinates, the other left to interested people to study.


Conclusion how to avoid being reverse engineering

See here, as a developer may feel no way to prevent others to do reverse engineering analysis, in fact, yes.

Blur your  Java  code is the first step: use  Proguard  .  It will  remove  all the  names  of  packages  ,  fields  and  methods  with random numbers, making analysis more difficult.  If you want to  do this analysis  of the  App  , starting from the  framework classes  .  Proguard is  not only used in fuzzy program code, but also can remove useless  resources  and  methods  .  Proguard is  very easy to use, I think  Pokémon Go  should be used in the future. 

There is also a way to reduce the  Java  program code, part of the function re% into  native libraries  , which will increase the difficulty of analysis, but the development is very convenient, and there are too many  Java  and  native in  series, will lead to Performance degradation

WE CAN Intercept at The Network Request Because at The  App  DID not use  Certificate pinning  .  The Using  Basic Android classes  or  OkHttp  IS Very the Common and the Easy.  But like Fuzzy Program code, IT CAN not resist at The Extreme Attackers  (  Because at The Certificate CAN Also BE Reverse Engineering  )  , But can lag them for some time.     

Finally, this article is quite basic to the analysis, we did not expose the secrets of any game, open cheating methods make the game unfair.  But for developers, you have to be careful to guard against professional hackers.

Here's a look at our findings

The           program code is not blurred, it is easy to reverse engineer the analysis.

·           We can rebuild executable project

·           Dependency management library could be better

·           Future without VR or Cardboard signs version    

·           May reduce Android version requirements  

·           We can read the  code associated  with the location / network / sensor and Pokémon Go Plus     

·           Easy to fetch network requests because of lack of certificate pinning 

·           Network requests is through protobuffers-RPC completed  

You can find our reverse engineering version of the program code:  Github  contact us:  Twitter

如社区发表内容存在侵权行为,您可以点击这里查看侵权投诉指引

标签: