PureScript FFI Best Practices

These are some patterns I use from day to day when writing PureScript FFI code, I call them best practices, but no need to take them too seriously, use your own judgement. To understand why a foreign import works or doesn't work, the best way is looking into the generated JavaScript code in the output folder directly.

Best practices

1. Use arrow functions to curry manually
2. No type class constraint in foreign import
3. Use Json to interoperate
4. Use Promise for async code

How to use this repo

src
└── Practice
    ├── ArrowFunction
    │   ├── Main.js
    │   ├── Main.purs
    │   └── README.md

Each practice lives in its own folder inside src/Practice, each folder contains three files. Main.purs is auto-extracted from README.md, so it's recommended to open Main.js and README.md side by side.

You can also run each practice separately

spago build
spago run -m Practice.ArrowFunction.Main
spago run -m Practice.Promise.Main

A Literate PureScript Setup

All src/**/Main.purs are extracted from src/**/README.md by sed.

Run find src/**/README.md | entr -s ./codegen.sh to regenerate Main.purs on README changes.

Some libraries following the best practices

• purescript-nonbili-postgres
• purescript-msgpack-msgpack

Support my open source works by hiring me, or using Hunch. Hunch is a website builder written in Haskell and PureScript.