insta-science

The insta-science Python project distribution provides two convenience console scripts to make bootstrapping science for use in Python projects easier:

• insta-science: This is a shim script that ensures science is installed and then forwards all supplied arguments to it. Instead of science, just use insta-science. You can configure the science version to use, where to find science binaries and where to install them via the [tool.insta-science] table in your pyproject.toml file.
• insta-science-util: This script provides utilities for managing science binaries. In particular, it supports downloading families of science binaries for various platforms for use in internal serving systems for offline or isolated installation.

This project is under active early development and APIs and configuration are likely to change rapidly in breaking ways until the 1.0 release.

Configuration

By default, insta-science downloads the latest science binary release appropriate for the current platform from the science GitHub Releases and caches it before executing for the 1st time. You can control aspects of this process using the [tool.insta-science] table in your pyproject.toml file. Available configuration options are detailed below:

Offline Use

There is full support for offline or firewalled science use with insta-science. You can seed a repository of science binaries by using the insta-science-util download command to download science binaries for one or more versions and one or more target platforms. The directory you download these binaries to will have the appropriate structure for insta-science to use if you serve up that directory using your method of choice at the configured base url. Note that file:// base URLs are supported.

Likewise, you can seed a repository of ptex binaries, scie-jump binaries and interpreter provider distributions by using the insta-science download {ptex,scie-jump,provider} ... family of commands and updating corresponding base_url options in your scie lift manifest.

Development

Development uses uv. Install as you best see fit.

With uv installed, running uv run dev-cmd is enough to get the tools insta-science uses installed and run against the codebase. This includes formatting code, linting code, performing type checks and then running tests.