coursebuilder/docs/quickstart.md

Concept

The concept behind coursebuilder is to store curricula descriptions in YAML files that can be versioned in a git repository. Unlike classic databases an observable and well defined versioning is paramount in these descriptions as they are the legal foundation for study and exam regulations.

The following pieces play together here:

• schema files, usually a schema.yaml
• mod files, usually something along the lines of mod.coursecode.yaml
• book files describing a whole regulation set and course global details
• some sort of transformation with coursebuilder into Markdown that is piped through pandoc in order to generate PDF, HTML and other representation from this code

schema files

Schema files are responsible to describe the used structures in a database. The following datatypes are supported:

• str a simple string, can be accompanied with a template
• enum a classic enum datatype with a fixed set of values
• num a numeric datatype
• multinum an array type with the possibility to spec each value
• multikey a key-value type with additional numeric data associated with each key instance

mod files (modules)

Modules describe a course in detail and implement an instance of the schema file. Especially strings and enums are translatable One of the plan is to use a validator to find inconsistencies automatically, like workloads that are not following the 30h = 1ECTS rule.

datatypes

str datatype

# this would reside in a schema field on top level
# a field of name 'id'
id: # name of the field 
    type: str # sets the datatype to str
    translatable: false # enforces the value is not translatable (default is true)
    label: { # label describes the meaning of the datatype in regards of the schema
        de: "Kürzel",   # translation of the label in German (de)
        en: "code" # translation of the label in English (en)
    }