This repository contains a collection of CMake scripts to help you embed Google's native WebRTC implementation inside your project as simple as this:
cmake_minimum_required(VERSION 3.9)
project(sample)
find_package(LibWebRTC REQUIRED)
include(${LIBWEBRTC_USE_FILE})
set(SOURCE_FILES main.cpp)
add_executable(sample ${SOURCE_FILES})
target_link_libraries(sample ${LIBWEBRTC_LIBRARIES})
- CMake 3.9 or later
- Python 2.7 (optional for Windows since it will use the interpreter located
inside the
depot_tools
installation)
- Required development packages:
# apt-get install build-essential libglib2.0-dev libgtk2.0-dev libxtst-dev \
libxss-dev libpci-dev libdbus-1-dev libgconf2-dev \
libgnome-keyring-dev libnss3-dev libasound2-dev libpulse-dev \
libudev-dev
- GCC & G++ 4.8 or later, for C++11 support
- OS X 10.11 or later
- Xcode 7.3.1 or later
-
Windows 7 x64 or later
-
Visual Studio 2015/2017
Make sure that you install the following components:
- Visual C++, which will select three sub-categories including MFC
- Universal Windows Apps Development Tools
- Tools (1.4.1) and Windows 10 SDK (10.0.14393)
-
Windows 10 SDK with Debugging Tools for Windows or Windows Driver Kit 10 installed in the same Windows 10 SDK installation directory.
Clone the repository and run script to fetch WebRTC sources. You can pass WebRTC version (branch) to script to fetch specific version.
git clone https://github.com/UltraCoderRU/libwebrtc.git
cd libwebrtc
python3 sync.py [WEBRTC_VERSION]
Create an output directory, browse inside it, then run CMake to configure project.
mkdir build
cd build
cmake -G Ninja -DCMAKE_BUILD_TYPE=<Debug/Release> -DCMAKE_INSTALL_PREFIX=<install_path> ..
After configuration build library using standard CMake commands.
cmake --build .
cmake --install .
The library will be located inside the lib
subdirectory of the <install_path>
.
The include
subdirectory will contain the header files.
CMake scripts will be placed inside the lib/cmake/LibWebRTC
subdirectory.
To import LibWebRTC into your CMake project use find_package
:
find_package(LibWebRTC REQUIRED)
include(${LIBWEBRTC_USE_FILE})
target_link_libraries(my-app ${LIBWEBRTC_LIBRARIES})
You should add library installation path to CMAKE_INSTALL_PREFIX
variable
while configuring your project:
cmake -DCMAKE_PREFIX_PATH=<install_path> ...
The library will be compiled and usable on the same host's platform and architecture. Here are some CMake flags which could be useful if you need to perform cross-compiling.
-
GN_EXTRA_ARGS
Add extra arguments to the
gn gen --args
parameter. -
NINJA_ARGS
Arguments to pass while executing the
ninja
command.
The following table displays the current state of this project, including supported platforms and architectures.
x86 | amd64 | arm | arm64 | |
Linux | - | ✔ | - | - |
---|---|---|---|---|
Windows | - | ✔ | - | - |
MacOS | - | - | - | - |
Android | - | - | - | - |
iOS | - | - | - | - |
Many thanks to Dr. Alex Gouaillard for being an excellent mentor for this project.
Everything started from his «Automating libwebrtc build with CMake» blog article, which was a great source of inspiration for me to create the easiest way to link the WebRTC library in any native project.