Abstract

The venerable Fortran programming language dates back to the 1950s, and to this day remains a cornerstone of scientific computing, underpinning subjects such as materials discovery, climate modelling, and fusion energy research. Yet its long history presents challenges: like geological strata, modern well-structured code rests atop software written to ancient standards and riddled with obsolescent programming techniques. In addition, the developer tooling ecosystem has significantly lagged behind more popular languages like C++ and Python, hindering modernisation efforts.

To address these challenges, we have developed Fortitude – a linter which detects bugprone code, style violations, and opportunities for modernisation, and in some cases is able to fix these issues automatically. Fortitude has developed rapidly from an in-house tool to one with a large and growing user base. By leveraging existing assets like Ruff, an open-source Python linter, and Tree-Sitter, a general-purpose language parser framework, Fortitude delivers high-performance, a familiar user interface, and robust Fortran parsing. This reuse of proven technologies has also accelerated development, allowing Fortitude to jump straight to the cutting-edge.

This talk will cover the development story of Fortitude, from its conception and design choices to its future roadmap. It will also demonstrate how Tree-Sitter can be used to rapidly develop static analysis tools.