BehaviorTree.CPP

BehaviorTree.CPP is a C++17 framework for creating Behavior Trees, designed for robotics, game AI, or as a replacement for Finite State Machines.

Key features include:

• Asynchronous Actions: Non-blocking actions are first-class citizens.
• Reactive Behaviors: Support for concurrent action execution (orthogonality).
• XML Scripting: Trees are defined using an XML-based domain-specific language and can be loaded at runtime, meaning tree morphology is not hard-coded.
• Extensibility: Custom TreeNodes can be statically linked or converted into plugins for runtime loading.
• Dataflow: A type-safe mechanism for data exchange between nodes.
• Logging/Profiling: Infrastructure for visualizing, recording, replaying, and analyzing state transitions.

The library is built upon several fundamental concepts:

• BT::PortInfo: A type-safe port system used to define inputs and outputs for nodes.
• BT::Expected: A result type used for robust error handling.
• BT::NodeStatus: Represents the execution states of a node (e.g., Success, Failure, Running).

To install cppzmq, you must first build and install libzmq. Follow these steps:

1. Build libzmq:

   • Clone libzmq from https://github.com/zeromq/libzmq.git.
   • Perform an out-of-source build using cmake and install it using make install.

2. Build cppzmq:

   • Clone cppzmq from https://github.com/zeromq/cppzmq.git.
   • Perform an out-of-source build using cmake.
   • You can skip building tests by using the flag -DCPPZMQ_BUILD_TESTS=OFF.
   • Install using make install.

# 1. Build libzmq
git clone https://github.com/zeromq/libzmq.git
cd libzmq
mkdir build
cd build
cmake ..
sudo make -j4 install

# 2. Build cppzmq
git clone https://github.com/zeromq/cppzmq.git
cd cppzmq
mkdir build
cd build
cmake ..
sudo make -j4 install

To use minicoro, include the header in exactly one .c file with the MINICORO_IMPL macro defined. In all other files, you can include minicoro.h as a standard header.

#define MINICORO_IMPL
#include "minicoro.h"

To use minitrace, include minitrace.c and minitrace.h directly in your project. You should add #include minitrace.h to a common header file.

To enable profiling, you must define the MTR_ENABLED macro globally during compilation (e.g., using the -DMTR_ENABLED flag in your compiler settings). Ensure you remove this definition in final release builds.

Use the provided CMake find package scripts to include cppzmq in your project. This will also include libzmq for you.

To link against the dynamic library, use cppzmq. To link against the static library, use cppzmq-static.

# find cppzmq wrapper, installed by make of cppzmq
find_package(cppzmq)
target_link_libraries(*Your Project Name* cppzmq)

# Or use static library to link
target_link_libraries(*Your Project Name* cppzmq-static)

For detailed usage, tutorials, and API references, use the following resources:

• Tutorials and General Documentation: https://www.behaviortree.dev/
• Auto-generated Doxygen: https://behaviortree.github.io/BehaviorTree.CPP/
• Community Support: Use the GitHub Discussions for forum-style help.

If you have dependencies like ZeroMQ and SQLite already installed on your system, you can build using standard CMake commands:

mkdir build_release
cmake -S . -B build_release
cmake --build build_release --parallel

You can use the vcpkg dependency manager to build and install the library:

git clone https://github.com/Microsoft/vcpkg.git
cd vcpkg
./bootstrap-vcpkg.sh
./vcpkg integrate install
./vcpkg install behaviortree-cpp

Minitrace produces JSON files compatible with Chrome's built-in trace viewer.

1. Open Google Chrome and navigate to chrome://tracing (or about:tracing).
2. Click the Open button.
3. Select your generated .json trace file.
4. Use the WASD keys to navigate the timeline and identify performance bottlenecks.