This project is an example of a minimal build system for Standard ML using smlnj as the compiler and ninja to orchestrate the build. Notable features are:
- Generates native executables by bootstrapping a heap2asm build (not included with the standard smlnj distribution).
- Uses CM to generate depfiles which ninja is then aware of (not supported by the standard tooling).
- This project comes with a working github CI setup (with a Linux runner), and should also build just fine on macOS.
To build, ensure SMLNJ_HOME is set appropriately, and then:
./configure.py
ninja