Modules​

Every EdgeDB database can contain several modules, each with a unique name. Modules can be used to organize large schemas.

module default {
  type BlogPost {
    required property title -> str;
    required link author -> auth::User;
  }
}
module auth {
  type User {
    required property email -> str;
  }
}

Modules can contain a wide range of schema elements: object types (equivalent to tables in SQL), custom scalar types, expression aliases, abstract links and properties, functions, and more. Each of these schema elements has a name; no two elements in the same module can share the same name.

You can split up your schemas however you see fit. Most users put their entire schema inside a single module called default.

Fully-qualified names​

When referencing schema objects in a different module, you must use a fully-qualified name of the form module_name::object_name. In the schema above, note how BlogPost.author points to auth::User. If User and BlogPost were in the same module the auth:: prefix wouldn’t be necessary.

Standard modules​

EdgeDB contains the following built-in modules which come pre-populated with useful types, functions, and operators. These are read-only modules. The contents of these modules are fully documented in Standard Library.

• std: standard types, functions and other elements of the standard library

• math: algebraic and statistical functions

• cal: local (non-timezone-aware) and relative date/time types and functions

• schema: types describing the introspection schema

• sys: system-wide entities, such as user roles and databases

• cfg: configuration and settings