chore(python): support uv dependency management in cgraph

[weijinglin]

How to integrate CGraph with uv

Add the following command inn your pyproject.toml file (configuration file in uv ecosystem) under [tool.uv.sources] section

pycgraph = { git = "https://github.com/ChunelFeng/CGraph.git", subdirectory = "python", rev = "branch_support_uv", marker = "sys_platform == 'linux'"  }

With this configuration, uv will download the source from GitHub on linuxOS, and fall back to PyPI on all other platforms.
After declaring the target source of pycgraph, you need declare the pycgraph as a dependency of your project.

[ChunelFeng]

can u show the result in your branch? Or give a example how to install PyCGraph in linux with this?

[weijinglin]

Of course, I have tested on my machine (ubuntu 24). Test on mac & windows is necessary. I will introduce the example in the following section.

uv init example-graph # init a new virtual environment

cd example-graph

uv add "pycgraph @ git+https://github.com/ChunelFeng/CGraph.git@main#subdirectory=python" --marker "sys_platform == 'linux'" # use this commond to add pycgraph as a dependency, and we can replace the repo url and branch configuration if this pr is merged

uv sync

then, we edit the main.py generated by uv

from PyCGraph import GPipeline, GNode, CStatus

class HelloCGraphNode(GNode):
    def run(self):
        print("Hello, PyCGraph.")
        return CStatus()

def tutorial_hello_cgraph():
    pipeline = GPipeline()
    pipeline.registerGElement(HelloCGraphNode())
    pipeline.process()


if __name__ == '__main__':
    tutorial_hello_cgraph()

At last, we run the main.py scripts to test basic functionality

expected output: Hello, PyCGraph.

[ChunelFeng]

thank you