Dissecting pkp-lib - The Core of Academic Publishing Systems

If you've ever submitted a paper to a university journal or searched for a peer-reviewed preprint, you've almost certainly encountered the Open Journal Systems platform. It's a classic of academic web. But few developers think about what holds such platforms together.

Behind the scenes of the Public Knowledge Project's academic software family is the PHP library pkp-lib. It contains shared logic for several products at once, including the OJS journal platform, OMP monographs, OPS preprints, and OHS metadata harvester.

What's Inside and Why It Was Done This Way

Instead of copying code between different projects, the team extracted the core functionality into a separate repository. The library is included as a submodule in the lib/pkp folder within each application.

The internal architecture handles the routine:

• Authorization, role management for reviewers and editors.
• Support for academic metadata standards OAI-PMH and Dublin Core.
• A hook engine for extending functionality.
• Common UI components and manuscript submission workflow handling.

It's interesting to look at the GitHub statistics. The repository has a modest 350 stars. But there are almost five hundred forks, and the number of open issues has exceeded a thousand. For open academic software, this is a normal picture. People come here not for the stars, but to solve real work tasks.

How Bug Tracking is Organized

The main README of the project looks restrained. There are no examples of terminal commands or instructions for installation via Composer. The reason is simple: the repository serves as a single dispatch point for errors across the entire application family.

If a scientific journal editor finds a bug in OJS 3.3, they create a ticket right in pkp-lib. Maintainers link the issue to a specific core release, after which the fix makes its way into all dependent platforms. For long-running products with a small development team, this approach helps keep the chaos under control.

Room for Code to Grow

The project has existed since 2009, so the codebase contains architectural layers from different eras of PHP development. Legacy code coexists with MVC components, and refactoring proceeds incrementally from release to release.

Studying the repository makes sense in several cases:

• You've been assigned to extend or customize a university journal running on OJS.
• You're writing a plugin for article export or citation database integration.
• You want to see how a shared core is organized for several specialized products.

Using pkp-lib as a framework for an arbitrary website is not advisable. The library has a narrow specialization for publishing workflows. But as an example of a domain-driven system with a long history, it provides interesting food for thought.