This is a fork of sysdream/Protod. Credits to the creators of Protod.

This code has successfully been used to extract .proto files from a binary that used the proto2 syntax.

However, it is untested on proto3.

Protod - Protobuf's metadata extractor

(c) 2012, Sysdream ([email protected])

Changes:

• Adapted to Python 3.10

• Code has been refactored to (hopefully) make it easier to understand

• New functionality that improves execution time, in my case by ~50x

  • NEW: Specify pattern to search for (instead of b'.proto', you might choose b'.proto\x12'. In my case, this avoided many "false positives")
  • NEW: Ignore certain filenames by passing a filename predicate to the ProtobufExtractor. For instance:

  def filename_predicate(filename: str):
      return "google/protobuf" not in filename
  
      [...]
      extractor = ProtobufExtractor(sys.argv[1], filename_predicate=filename_predicate)
      extractor.extract(byte_pattern=b".proto\x12")

To extract every .proto file from a given executable:

python src/protod.py somebinary

• To change what filenames are included, modify the filename_predicate function in src/protod3.py.
• To change the byte_pattern (most permissive to catch all proto files: b'.proto'), edit the byte_pattern parameter that is passed to ProtobufExtractor::extract. b'.proto\x12' worked well for my use case.

Protod is a tool able to extract Google's protobuf metadata from any binary file. This version has been designed to cover every file format.

The goal of this tool is to recover serialized protobuf's metadata inserted at compilation time inside an executable, and to make it available as .proto file, ready to compile with protoc (protobuf's compiler).

For further information on Google's protobuf library, please see:

https://developers.google.com/protocol-buffers/docs/overview