A multi purpose node editor for flow-based programming.

The purpose of this node editor is to connect functional building blocks and parameterize them in order to form a graph which can then be executed using the execution subsystem.

For running the graph created with the node editor, some sort of flow-based programming library is needed.

If you want, you could swap the execution subsystem shipped in the Releases by your own system. It just needs to understand the node editor's output format of the graph.

At the present moment the nodes available in the node editor is limited to only a couple of image processing nodes. You can add your own node types by adding a JSON File to a subfolder of apps\execution_node_editor\execution_subsystem\node_type_definitions containing your node attributes. Here is one example:

{
    "node_type": "GaussianBlurNode",
    "input_ports": [
      {
        "port_name": "image",
        "data_type": "image_t"
      }
    ],
    "output_ports": [
      {
        "port_name": "blurred",
        "data_type": "image_t"
      }
    ],
    "default_settings": 
    {
      "sigma": 1.0
    }
}

The node editor creates two files

• The Scene with the file extension .nes saves the entire scene containing nodes, edges, node settings, viewport settings and user interface settings. It is designed to be loaded again by the node editor.
• The Graph with the file extension .graph.json saves only the nodes, edges and node settings. It is designed to be an exchange format for the execution subsystem and running the graph.

This software is licensed under MIT License.

• Anaconda or Miniconda. Get it from: https://docs.conda.io/projects/conda/en/latest/user-guide/install/index.html
• On Windows only you need Inno Setup for creating the Installer. Get it from: https://jrsoftware.org/isdl.php. Install it after you downloaded it.

Create conda environment using

conda env create -f environment.yml

This will create a new conda environment called editor. Now activate the environment using

conda activate editor

Install randomname

Randomname is used to create random file names by the editor. This fork of the original package is used.

You need to take the fork because the original version can not be installed properly using the PyInstaller.

Clone the repository:

git clone [email protected]:beyse/randomname.git
cd randomname

Make sure that the editor environment is activated and run:

python setup.py install

Now run the packaging script in a Terminal with the conda environment editor activated.

For Windows run:

.\pack.bat

For Linux or macOS run:

./pack-ubuntu.sh

You should now see the dist folder. It contains the executable file ExecutionNodeEditor or ExecutionNodeEditor.exe on Windows alongside with the two folders assets and execution_subsystem.

• On Windows it contains the Installer Setup ExecutionNodeEditor.exe.
• On Linux or macOS it contains a ZIP file called ExecutionNodeEditor which can be shipped.

• This node editor is based on pyqt-node-editor by Pavel Křupala. Visit the original repository on GitLab.

• The code contains a modified version of PyQJsonModel

If you would like to contribute please read this Contribution Guide

Feel free to contact me if you have any questions: [email protected]