Two years ago I started working on a new formatter for Haskell source code. My idea was to take advantage of the parser of GHC itself, which finally became available as a library around that time. Soon Tweag supported the initiative and the project became Ormolu. It was announced in May, right before ZuriHac 2019. Many people kindly helped me with it during the Hackathon. We went on to release the first version in October 2019.
The use of the GHC parser and a solid approach to testing won Ormolu the reputation of a dependable tool. More and more industrial users choose it as their Haskell formatter. In summer 2020 a rather big company had decided that they wanted to format their Haskell code with Ormolu, which resulted in a three-month full-time contract. A new level of quality was reached.
I think it is cool to be paid to work full-time on a project like this. If you concur, we have good news for you! Tweag currently has an opening for an intern who would like to work on the formatter. This is a project for someone who would enjoy iteratively improving a popular Haskell tool and learn about GHC AST, its parser, and perhaps a bit of Nix.
There are various issues affecting Ormolu, and fixing these would have a positive impact on the user experience. The internship would address them in severity order:
- Upgrading the GHC parser. Presently, Ormolu uses
ghc-lib-parser. GHC 9.0 fixes some long-standing issues, and we can take advantage of that by switching to
- Some more bugs of varying difficulty.
- Stylistic changes.
Please let us know if you are interested! Include a cover letter describing your Haskell experience; familiarity with GHC AST is a plus. We will collect applications till Tuesday, June 1, 2021. The internship can start any time after the offer is made, subject to mutual availability. Internships typically last 12 weeks, although the duration may be adjusted if necessary. If you have any questions, feel free to email me.