Skip to main content

Create a Domain Model

Use a Database Schema to Communicate Your Domain Model​

A ZefDB schema is different from schemas in other databases. There are two distinct parts which we distinguish:

  • constraints: a set of logical constraints on any valid DB state. These can be simple local constraints (as in relational DBs), non-local constraints or anything you can express in a predicate function
  • blueprint: describe the shape of your data. If data exists in a certain shape, this shape is guaranteed to show up in the blueprint. 
[  
    ET.Movie[  
        'title':           String,  
        'year_of_release': Year,  
        'actors':          List[ET.Person],  
        'directors':       List[ET.Person],  
        'writers':         List[ET.Person],  
        'genres':          Set[ET.Genre],    # or tagged?  
        'age_rating':      ET.AgeRating,  
    ],  
    ET.Person[  
        'first_name': String,  
        'last_name': String,  
        'year_of_birth': Year,  
        'country_of_birth': ET.Country,  
        ...               # there may be more fields  
    ],  
    ET.Country[  
        'name': String,  
        'population': Int,  
    ],  
    ET.Genre[  
        'name': String,  
    ],  
]

Note that this is a valid Python expression: a list of types.
A blueprint describes the structural shape of your knowledge graph on a type level.

Optional Fields​

General Optional Fields​

ET.Movie[  
    'title': String,  
    ...  
]

using ellipses indicates that all other possible fields are optional. title is required, but if any additional field is added, it will still be part of the type.

Individual Optional Fields​

One can add specific optional fields though. 

ET.Movie[  
    'first_name': String     : Optional,  
    'last_name': String,  
    'year_of_birth': Int     : Optional,  
]

This means: if these fields are present, they MUST be of the specified type.

Alternative syntax: 

ET.Movie[  
    Optional['first_name']: String,  
    'last_name': String,  
    Optional['year_of_birth']: Int,  
]  
ET.Movie[  
    'first_name?': String,  
    'last_name': String,  
    'year_of_birth?': Int,  
]