datonic / seamless

📦 Oxidizing Frictionless Standards

📦 Seamless

A minimal CLI to work with Frictionless Data Packages.

Installation

Download the latest release and put it seamless in your PATH. Alternatively, you can install directly from GitHub with cargo:

cargo install --git https://github.com/datonic/seamless

Usage

The following commands are available:

seamless pull <path> <path>
seamless query <path> <query>
seamless validate <path>

License

MIT License

seamless's Issues

Ideas for first version

As a first approach, get the following commands running:

• sd¹ pull example.org/datapackage.json
• sd query example.org/datapackage.json

---

¹ sd as Seamless Data

Create a Package Registry

Basically a GitHub repository with some famous/useful datasets. Idea from "Writing a Package Manager".

Understand platform specific URIs

A sane default might be to rely on the information URIs give us and be smart about it.

For example, say we wan't to use a HuggingFace dataset, davidwisdom/fake_railroad_company. Seamless should know how to get and parse the URI hf://davidwisdom/fake_railroad_company. Same might happen with kaggle or other platform with a clear way to identify the dataset.

This is a different mode that pointing to a datapackage.yaml file though and might be confusing.