MLscript

This is a pre-release of MLscript.

Quick Start

To install MLscript, you can use npm:

npm i mlscript

The newest version is recommended: there may be lots of bugs in the previous versions.

To watch the project directory, you can use TypeScript(recommended)/JavaScript:

import mlscript from "mlscript"

// filename: string, workDir: string, outputDir: string, tsconfig: string, commonJS: boolean, expectTypeError: boolean
mlscript.watch("./mlscript/Main.mls", "./", "./js/", "./tsconfig.json", false, true)

The path must start with ./, or it will be treated as node_modules paths.

If you need to interoperate with TypeScript, you must provide the path of tsconfig.json file. Otherwise, it is optional.

You can choose to use CommonJS or ES modules(recommended).

Since Mlscript is still under development, type inferences are not fully supported. If there is a type error but it is correct, please set the expectTypeError flag to true. To get the type error messages, please check the corresponding .mlsi files in .interfaces.

For more details, you can check the demo here.

Project Structure

The MLscript uses git to track the file changes. Please make sure your project locates in a git repo.

Though not necessary, we still recommend you organize the project in this structure:

• .interfaces: temporary interface files generated by MLscript.
• mlscript: directory for MLscript files.
• node_modules: third-party libraries in TypeScript.
• ts: directory for TypeScript files.
• configuration files: package.json and tsconfig.json should locate in the root directory.