lukaszgemborowski/nanopbpp

C++ helpers for nanopb library

nanopb++

Lightweight C++ wrapper(s) library for nanopb implementation. The goal is to simplify some common tasks using C++ and metaprogramming wherever possible.

Original nanopb implementation can be found here: https://github.com/nanopb/nanopb and it's homepage is here: http://jpa.kapsi.fi/nanopb/

Idea

Library is simplifying some common tasks using modern C++ features and it is also introducing more type safety to your project. To achieve it nanopb++ is introducing something which I personally like to call "metadata". This is description of your messages in C++ with some helper templates. It's most painfull part of using nanopb++ but it's done only once for a set of messages. In some future this task may be automated but for now it must be done manually. Example of such description for this proto file:

syntax = "proto2";
import "nanopb.proto";

message IntegerContainer {
	required int32 a = 1;
	optional int32 b = 2;
}

message FloatContainer {
	required float c = 1;
}

message Extendable {
	extensions 10 to 100;
}

extend Extendable {
	optional IntegerContainer field_a = 11;
	optional FloatContainer field_b = 12;
}

would be something like this:

static nanopbpp::message_metadata_set<
	nanopbpp::message_metadata<IntegerContainer>,
	nanopbpp::message_metadata<FloatContainer>,
	nanopbpp::message_metadata<Extendable,
		nanopbpp::extension_metadata_set<
			nanopbpp::extension_metadata<field_a_tag, IntegerContainer>,
			nanopbpp::extension_metadata<field_b_tag, FloatContainer>
		>
	>
> messages_metadata(
	{ IntegerContainer_fields },
	{ FloatContainer_fields },
	{ Extendable_fields, {field_a, field_b} }
);

it's creating a relationship between message type (like IntegerContainer) and it's pb_field_t description. Additionally it can describe extensions of the message. Now if you have such meta-description you can use it with nanopb++ helper functions such as encode and decode (extract from test cases):

using namespace nanopbpp;

IntegerContainer source = {0}, destination = {0};

// prepare data to be encoded in protobuf, basically
// you need to fill standard nanopb generated structure
source.a = 10;
source.has_b = true;
source.b = 20;

std::array<uint8_t, 64> buffer = {0};

// encode message to buffer
ASSERT_TRUE(encode(messages_metadata, buffer.begin(), buffer.end(), source));

// decode message from buffer
ASSERT_TRUE(decode(messages_metadata, buffer.begin(), buffer.end(), destination));

ASSERT_EQ(source.a, destination.a);
ASSERT_EQ(source.b, destination.b);
ASSERT_EQ(source.has_b, destination.has_b);

simple, huh? I'm trying to keep all the basic scenarios as test cases in scenarios.cc file.

Current status

"work in progress" with very limited functionality and a small set of unit tests. To see use-cases please see tests directory.

Building

as for now the library contains only C++ headers and it does not require anything special, only compiler with C++11 support. If you want to run tests than you need some additional things:

• cmake >= 3.2
• protoc
• gtest library development files

additionally you need to checkout nanopb submodule in the project (git submodule update --init nanopb) and run make in nanopb/generator/proto directory. After that you can create some temporary build directory, run cmake and then make. Tests will be build for you, test executable name is simply "tests". You can run it now with ./tests command.