jelmer / upstream-ontologist

discover information about upstream projects
GNU General Public License v2.0
14 stars 2 forks source link
ontology upstream

Upstream Ontologist

The upstream ontologist provides a common interface for finding metadata about upstream software projects.

It will gather information from any sources available, prioritize data that it has higher confidence in as well as report the confidence for each of the bits of metadata.

The ontologist originated in Debian and the currently reported metadata fields are loosely based on DEP-12, but it is meant to be distribution-agnostic.

Provided Fields

Standard fields:

Extensions for upstream-ontologist, not defined in DEP-12:

Supported Data Sources

At the moment, the ontologist can read metadata from the following upstream data sources:

It will also scan README and INSTALL for possible upstream repository URLs (and will attempt to verify that those match the local repository).

In addition to local files, it can also consult external directories using their APIs:

Example Usage

The easiest way to use the upstream ontologist is by invoking the guess-upstream-metadata command in a software project:

$ guess-upstream-metadata ~/src/dulwich
Security-MD: https://github.com/dulwich/dulwich/tree/HEAD/SECURITY.md
Name: dulwich
Version: 0.20.15
Bug-Database: https://github.com/dulwich/dulwich/issues
Repository: https://www.dulwich.io/code/
Summary: Python Git Library
Bug-Submit: https://github.com/dulwich/dulwich/issues/new

Alternatively, there is a Python API as part of the upstream_ontologist Python package. There are also autocodemeta and autodoap commands that can generate output in the codemeta and DOAP formats, respectively.