Browser-based data tools using Pyodide

[MatMoore]

Idea description

Pyodide is a tool that lets you compile python code into web assembly and run it in a webpage, without needing a server.

I think a webpage using this would be great for quick, easy to use tools, for example:

• Using MOJAP metadata to convert between schema formats
• Infer a schema from a dataset, to help with manually authoring a schema
• Lint schemas

It supports Pandas, and PyArrow will be available in the next release

Why should we do this?

• Static web applications are an easy way of getting applications to users as they don't have to install or sign up for anything
• There is no need for data to leave the user's browser, so it is easy to secure
• No backend service is required, so it's easy to deploy (e.g. on github pages)
• Learn more about the frontend and emerging technology

Definition of done?

1. we have an HTML webpage
2. it does something interesting with data
3. all the data processing is coded in python 🐍

What skills do we need?

• python
• javascript (some)