A simplistic web-based GUI for visualizing and constructing modular data-processing pipelines. It provides a simple python interface for backend implementation. earlPipeline can become quite handy when one needs a GUI for some already existing data-processing library.

The package consists of three parts:

• frontend: web-based GUI, implemented in JavaScript, powered by Ember framework
• web server: Python web server, built with Tornado
• example backend: example backend implementation (simple calculator)

Python part of the package depends upon:

• Python 2.x
• Tornado
• logutils
• bidict >= 0.1.2

All the JS libraries are included in the distribution.

To install the library, simply run smth like::

$ sudo python2 setup.py install

To try it out, create a file server.py somewhere, with the following contents:

from earlpipeline import server
from earlpipeline.backends import calculator

if __name__ == '__main__':
    server.set_backend(calculator)
    server.run(pipeline_folder=".")

Then, run the script::

$ python2 server.py

If everything went well, you should now be able to navigate to http://localhost:5000, click "New Pipeline", and play around with GUI. You might also have a look at earlpipeline/backends/calculator.py to see how backends are implemented.

This is how your pipeline might look like: