hex-flow-core 0.0.1a5

HEXFELLOW hex-flow Core

HEX FLOW CORE

      

---

📖 Overview

What is hex_flow_core

hex_flow_core is the Python core library for the HEXFELLOW hex-flow framework. It provides Zenoh-based publish/subscribe node abstractions and launch configuration utilities, enabling multi-node orchestration for HEXFELLOW robots, sensors, and simulators.

What problem it solves

• Pub/Sub communication: Nodes publish and subscribe to topics over Eclipse Zenoh, supporting high-throughput binary payloads (images, commands, sensor data).
• Topic remapping: Each node can remap logical topic names to actual Zenoh keys via environment variables, enabling flexible deployment without code changes.
• Launch orchestration: LaunchConfig and NodeConfig generate YAML launch files consumed by hexflow run, managing multiple nodes, build steps, environment variables, and router settings from a single Python script.

Target users

• Engineers integrating HEXFELLOW robots into their systems.
• Researchers running experiments with HEXFELLOW robots.

📦 Installation

Requirements

• Python >= 3.10
• OS: Ubuntu (or other Linux)
• Core dependencies:
  • eclipse-zenoh >= 1.9.0
  • envlog == 1.0.0
  • hex_util_runtime >= 0.0.0, < 0.1.0
  • pyyaml >= 6.0

Install hex-flow-cli

For Ubuntu or any Debian-based system, install Zenoh and hex-flow CLI:

curl -L https://download.eclipse.org/zenoh/debian-repo/zenoh-public-key | sudo gpg --dearmor --yes --output /etc/apt/keyrings/zenoh-public-key.gpg
echo "deb [signed-by=/etc/apt/keyrings/zenoh-public-key.gpg] https://download.eclipse.org/zenoh/debian-repo/ /" | sudo tee -a /etc/apt/sources.list > /dev/null
sudo apt update
sudo apt install zenoh curl
curl -fsSL https://raw.githubusercontent.com/hexfellow/hex-flow/main/install.sh | sh

For other systems, please install zenohd yourself, then run the install script.

Install hex-flow-core from PyPI

pip install hex_flow_core

📑 API

NodeCallback

Low-level base class wrapping a Zenoh session with publisher/subscriber management, topic remapping, and structured logging.

from hex_flow_core import NodeCallback

node = NodeCallback("my_node")
node.start()

node.create_pub("my/topic")
node.pub("my/topic", b"hello")

node.create_sub("other/topic", lambda sample: print(sample.payload.to_bytes()))

while node.is_working():
    time.sleep(1)

node.stop()

Environment variables read at init:

Variable	Description	Default
HEX_FLOW_NODE_NAME	Overrides the name argument	constructor arg
HEX_FLOW_REMAP	JSON dict mapping logical topics to actual Zenoh keys	{}
RUST_LOG	Log level passed to envlog	"info"

Node

Higher-level subclass of NodeCallback that automatically buffers received messages in bounded deques, providing a polling-style get() interface.

from hex_flow_core import Node

node = Node("my_node")
node.start()

node.create_sub("sensor/data", maxlen=20)

data = node.get("sensor/data", latest=True)
if data is not None:
    print(f"received {len(data)} bytes")

node.stop()

NodeConfig

Describes a single node for launch: name, build/run commands, required flag, topic remaps, and environment variables.

from hex_flow_core import NodeConfig

cfg = NodeConfig(
    name="cam_publisher",
    run_cmd="python cam_pub.py",
    build_cmd="pip install -e .",
    required=True,
    remap_dict={"camera/rgb": "robot_a/camera/rgb"},
    env_dict={"RUST_LOG": "debug"},
)

LaunchConfig

Aggregates multiple NodeConfig entries and exports a YAML launch file for hexflow run.

from hex_flow_core import LaunchConfig, NodeConfig

launch = LaunchConfig(
    local_only=True,
    enable_tui=False,
    log_to_file=True,
    save_path="/tmp/my.launch.yml",
)

launch.set_node("cam", NodeConfig(run_cmd="python cam.py"))
launch.set_node("ctrl", NodeConfig(run_cmd="python ctrl.py"))

path = launch.export()
print(path)  # /tmp/my.launch.yml

The exported YAML can be used with:

hexflow run /tmp/my.launch.yml

💡 Example

We provide examples to test the installation. After installation, run:

git clone https://github.com/hexfellow/hex_flow_core.git
cd hex_flow_core
hexflow run examples/test.launch.py

This launches 8 command publishers (~1000 Hz each), 6 image publishers (~50 Hz each), and 1 subscriber that validates payloads and logs throughput.

Note: Examples require numpy (pip install numpy).

📄 License

Apache License 2.0. See LICENSE.

🌟 Star History

👥 Contributors