A fast, modular library to generate landscapes.

Changelog

See CHANGELOG.md to lean about the novelties of each version.

Use

Learning how the library is structured goes a long way in understanding how to operate the classes in this library. As such, you should read the section Extend the algorithms to learn more.

Every object is first initialized, then called upon. This allows a preparation step if needed, for the algorithms to bootstrap themselves. Generators and filters output a Terrain object, and Filters and Readers take in a Terrain object, so you can chain filters to achieve the desired effect.

The full documentation is available at https://solarliner.gitlab.io/terrainlib.

Development install

This project uses Pipenv to manage dev and production dependencies. Therefore, virtual environment creation and dependency installation is all done with a single command:

pipenv install --dev

Extend the algorithms

There are 3 types of classes, typically; generators, filters and readers. The names should be straightforward enough, but here are descriptions about them:

• Generators generate terrain out of input parameters. This allows terrain
  to be created without any prior material.
• Filters act on the terrain by applying transformations. They take a
  terrain (and some inputs) and output another terrain, transformed.
• Readers Read the terrain and output another type of object entirely. Use
  readers to save an image of your terrain, a 3D mesh, etc.

Each kind has its own base class, that is nothing more than an abstract __call__ function.

Any of these follow a two-step process:

1. The generator is initialized with input parameters. This allows setting up
   internal parameters, bootstrap routines, etc. but does not generate the
   terrain. This is done from the __init__ function.
2. The generator is then called through __call__ in order to actually process
   data in and/or out. The __call__ function should return a Terrain object, or,
   in the case of a reader, the resulting object.

All predefined algorithms in this library follow this convention, so see those for examples of implementation.

License

This project is licensed under a GNU LGPLv3 license. Non open-source, commercial projects are welcomed, but should ask for explicit licensing agreement. See the license file for more information.