@dougrich/grammar

Theoretical Structure

is and has structure.

An object is a set of hierarchical terms: for example, a male human name might be a name/human/male, while an english town might be a name/town/english. Asking for a name with no further hinting might result in either a male human name or an english town. This is an example: in a racially segregated fantasy world like Lord of the Rings, it might be better structured as name/human/town.

An object has a set of named fields. For example, a town/english might have the fields:

• Name, which is a name/town/english
• Mayor, which is a person
• Population, which is a population/town

Operations

generate generates the entire tree - essentially a generate and a hydrate

Contributing

Tests are critical. Write a test before you write a feature. Write a test after you write a feature. Write a test during the feature. Write lots of tests.

Stories

As someone who authors content for the grammar:

• [ ] I want a simple authoring language
• [ ] I want it broken up into multiple files for legibility
• [ ] I want it to be easy to add or remove content from
• [x] I want discrete fields (i.e. oneOf)
• [ ] I want variable count of fields (i.e. someOf)
• [ ] I want continuous fields (i.e. hue between 56 and 70)
• [ ] I want discrete numeric fields (i.e. 2d6)

As someone who uses the grammar to generate content:

• [x] I want to hit a button and get the entire tree
• [x] I want to hit a button and get the top level fields, then manually expand the rest of the tree
• [ ] I want to use a 'seed' to guarantee consistent results across multiple re-builds
• [x] I want to re-roll a specific field (this will invalidate the seed)
• [ ] I want to be able to download and view in a convenient format

As someone who uses this as a library:

• [x] I want to be able to use either a file system or a database as the underlying storage mechanism
• [ ] I want to be able to plugin my own operations (think graph-expand) that utilize the grammar language
• [x] I want it to be isolatend and parallelizable so that I can run multiple queries at once