slangwidge v0.3.0

slangwidge

A scripting language for NPCs. Built with OCaml/Reason and inspired by zzt-oop from the classic ZZT by Tim Sweeney.

Usage

For now, check the tests for examples on how to use the parser. It's used for parsing a script into a list of commands for a character to follow, so generally speaking, it takes a string and spits out a list of commands. Shocking, I know.

Development

The process of modifying and testing the parser is described below.

Dependencies

Developers will need the following dependencies to make and test changes to the parser:

• opam
• corebuild
• bucklescript
• npm
• yarn

Process

Development is a three part process: compiling the main project, compiling the parser, and running tests. Each described below.

Compiling the main project

This is a BuckleScript project that follows the standard build process. To compile the project in watch mode, run:

yarn start

Changes detected in the src/ folder will then trigger recompilation.

Compiling the parsers

The parser/ folder contains the files that must be edited to make changes to the parser. This includes lexer.mll, parser.mly, and ast.ml. See here for an explanation of how lexers and parsers work in OCaml.

After changing these files, run

yarn run prepare

This invokes corebuild to generate .ml files from the lexer and parser, and moves them into the src/ folder where they are compiled with bsb along with the rest of the Reason project.

Testing

This project uses bs-jest as a test framework. Tests are in the __tests__/ folder. New tests should be added for any new abilities added to the parser.