Program to extract protobufs compiled for C#
Branch: master
Clone or download
Type Name Latest commit message Commit time
Failed to load latest commit information.
mono-extractor Fixed missing reference to GoogleV1Inspector.cs within mono project Jul 23, 2018
.editorconfig Updated editorconfig for python files Jun 19, 2017
.gitignore Ignore built assemblies Jul 12, 2017 Add LICENSE,, and editorconfig files Mar 28, 2016
LICENSE Appended licences for binaries distributed with mono extractor project Jul 14, 2017 Fixed anchor link, example command and added link to vs2017 Aug 28, 2017
hsproto-packaging.ini Extended google decompilation support Jul 5, 2018 Write syntax attribute when decoding from bin Jul 15, 2018


A C# program to extract Protocol Buffer definitions compiled with Google Protos or SilentOrbit.

The file is discussed below.


There are 2 versions of the extractor project. Both projects have the same purpose, but target a different framework each.

  • Extractor; Targets .Net Core. See 1.
  • Mono-Extractor; Targets Mono. See 2.


Compile and run the Extractor project!

You can get the .Net Core compilation tools through 2 channels;

Download the (CLI) binaries by downloading directly from the site. When using the dotnet CLI to build the project, make sure to restore dependant packages first!


Update your Visual Studio 2017 installation to include the .Net Core cross-platform development, which can be found under section Other toolsets.

Using Visual Studio 2017 is recommended because it handles dependancies and builds the project correctly for YOU. There is a community version which is free to install and use.


Compile and run the Mono-Extractor project!

You can download the Mono tools from their site. Their msbuild executable should automatically restore dependant packages.


Compile the program. Give it the library files you want to decompile and if you want proto2 or proto3 syntax.

The program will do the following actions automatically:

  • resolve circular dependancies;
  • resolve name collisions;
  • generates proto2 syntax output.

Basic usage example:

proto-extractor --libPath "%HS_LOCATION%/Hearthstone_Data/Managed" 
--outPath "./proto-out" 

Extended usage example:

This command is used to generate protobuffer files for HS proto repository

proto-extractor --libPath "%HS_LOCATION%/Hearthstone_Data/Managed" 
--outPath "./proto-out" --proto2 --automatic-packaging --manual-package-file "%REPO%/hscode-proto-packaging.ini"

The section Options explains all parameters. The executable will write parameter parsing errors on standard outstream.



The option --proto3 will use the proto3 compiler to generate .proto files with protobuffer 3 syntax.

Defaults to False.

Resolve circular dependancies

The option --resolve-circular-dependancies will run a processor object that detects and solves circular dependancies for you. Both circular dependancies between types and namespaces are detected.

Defaults to True.

Resolve name collisions

The option --resolve-name-collisions will run a processor object that detects and solves all kinds of name collisions for you. See the processor code for more details.

Defaults to True.

Automatic packaging

The option --automatic-packaging will try to group namespaces under the same namespace if their names show similarities. The used algorithm is longest substring matching, with anchorpoint at the beginning of the string. Half-words are cut to the nearest namespace component.

Defaults to False.

Manual packaging

The option --manual-package-file "PATH-TO-INI-FILE" can be used manually move content of namespaces or specific types to other/new namespaces. See the file hscode-proto-packaging.ini in the root of the repo for examples.

It's important to keep the order of processing algorithms in mind! We shouldn't try to manually compensate the behaviour of the automatic dependancy resolver. This has to do with trying to keep a consistent layout of outputted protobuffer files regarding the absolute dependancy on the source material.

Defaults to "" (empty string) -> nothing will happen.

Order of processing algorithms

The execution order of processing algorithms is always as follows:

  1. Manual packaging of namespaces;
  2. Resolve circular dependancies;
  3. Automatic packaging of namespaces;
  4. Resolving name collisions.

Binary proto extraction

Use the file!

At the root of the repository you'll find a python3 script that extracts protobuffer files from arbitrary binary files.

This only works on proto files which have been compiled with the Google Protobuffer compiler tool (protoc)! Use the proto-extractor project (see above) for hearthstone related proto files.


The only requirement for this script is the Google Protobuffer package which you can install through pip. A quick setup guide is given below.

  1. Install virtualenv; pip3 install virtualenv
  2. Create a virtual environment inside the CWD; py -3 -m virtualenv ./.env
  3. Activate the environment; ./env/Scripts/activate
  4. Install protobuf package; pip install protobuf


  1. Activate the venv which was setup in the previous section.
  2. Run the proto extraction script; python -o [out-dir] [inputfile [inputfiles ..]]
  3. Extracted proto files can be found inside [out-dir], the script will output any found, valid filename.


Proto-extractor is licensed under the terms of the MIT license. The full license text is available in the LICENSE file.


proto-extractor is a HearthSim project. All development happens on our IRC channel #hearthsim on Freenode.

Contributions are welcome. Make sure to read through the first.