GSoC NIU Projects 2025: datashuttle#

If you are interested in any of these datashuttle projects, get in touch! Feel free to open a new topic on our Zulip GSoC channel and ask the community.

Our working language is English.

Allow Google Drive or AWS as remote storage

In systems neuroscience, a lack of standardisation in data organisation schemes creates a barrier to data-sharing and collaboration. datashuttle provides a Python API and terminal user interface (TUI) to allow researchers to create, validate and transfer folders in a standardised way.

Projects can be transferred between computer systems in datashuttle via SSH or mounting drives. We would like to support transfer to Google Drive or and Amazon Web Services (AWS) buckets. Under the hood datashuttle uses RClone, which already implements transfer to Google Drive and AWS.

Deliverables

  • Extend datashuttle functionality to permit transfer between a local filesystem and Google Drive or AWS. This will involve exposing new functionality within the Python API as well as the terminal user interface.

  • Tests to cover any added functionality.

  • Documentation for the new functionality.

  • Duration

Medium (~175 hours)

Difficulty

This project is well-suited for a beginner contributor to open source.

Required skills Experience with Python

Nice-to-haves Experience with network data transfers (e.g. rsync, rclone, Google Drive, AWS)

Potential mentors

Further reading

datashuttle and RClone websites.

Deploy datashuttle as installable package

Currently, datashuttle requires conda to use. It would be great if the terminal-user-interface it could be deployed as an executable installation across Windows, macOS and Linux. This would make it accessible for those without coding experience.

Unfortunately, textual renders poorly in native system terminals, and so would likely require also deploying a terminal to run in. There are existing implementations packaging textual in cross-platform executables but no standard way to proceed.

This project would involve researching approaches, and implementing the most robust and maintainable approach for datashuttle.

Deliverables

  • Research approaches for cross-platform distribution of a datashuttle executable

  • Assess how robust and maintainable these are

  • Implement the best approach for datashuttle

  • Document and test in our GitHub CI

  • Duration

Large (~350 hours)

Difficulty

The project will require research deep into the python packaging ecosystem and mechanics, and so would best suit someone who is already experienced with Python. Knowledge of Python packaging is not required, but a good understanding of Python to use as a starting point is.

Required skills Experience with Python

Nice-to-haves Experience with created executable Python distributions (e.g. with PyInstaller).

Potential mentors

Further reading

datashuttle and textual.

This Page