# Glossary
The documentation tries to stick with certain terminology. This glossary is provided to help with understanding those terms.
Revisions Welcome!
We don't claim this glossary to represent ultimate truth. This glossary is open for discussion and correction.
The glossary isn't sorted alphabetically to help with reading it top to bottom, though some aspects might be given redundantly to understand every term individually.
# Model
A model combines a defined set of structured information with associated behaviour and applies a somewhat unique label to this combination. It is used to describe a certain type of item to be handled in code. Models can be derived by means of extending an existing model by adding or replacing information and/or behaviour resulting in a different, but related model.
For every model there may be a set of instances sharing that model as a description of their common structure and behaviour.
A model in ODM is very similar to a class in object-oriented programming.
Example
When creating an address book application it is going to handle records each describing a particular person. In this scenario person is a model.
It might define to have a name and some mail address for contacting this particular person. This is the structured information common to all the records and thus also known as a person's data.
The model will expose a method for sending mail to a particular person. This is the behaviour associated with a person's data.
The address book might support different kinds of persons, such as friends and colleagues. Either kind may be considered a derived model and it might adjust information structure and associated behaviour, e.g. adding birthday per friend to what is managed by common person model.
# Schema
When using a model's API the definition of properties and behaviour per model is exposed as the model's schema.
# Properties
In context of a model properties are used to describe the common structure of information found in instances of that model.
A definition of a property includes
- its name for identifying related information in context of instances of the model,
- its type of information used in every instance,
- constraints to be applied to detect valid/invalid information and
- indices used to improve performance on searching for/sorting by this property.
Example
One property of model person in the address book application will declare to have a last name (which is the identifying name of some information) as a string of characters (that's the type of information) and that providing a last name is required (which is a constraint applied to this information).
# Computed Properties
A model might define virtual properties that won't be stored persistently but will derive their information from other information of an instance on demand.
Example
When having actual properties last name and first name there might be a computed property full name that is combining the former two properties.
# Methods
A method is a piece of code defined in context of a model. However, there are two types of methods with regards to context when running:
A static or model-related method is running bound to the model, thus basically incapable of processing a particular instance of that model.
An instance method is running bound to a particular instance of the model it was defined in. This method is capable of processing the particular instance as well as accessing its model and use static methods defined there.
Example
Searching a person by its name is an operation to be run in context of whole model rather than a single instance of it. Thus this method is a static method of model person.
Sending mail to a particular person requires information of that particular person as context and thus is defined as an instance method.
# Item or Instance
A single set of values complying with the structure of a particular model may be called an item or instance of that model.
In code, instances aren't just complying with one of the models, but have a strong relationship with the model instead for the model's API is used to actually access and manage some of its instances.
Example
The address book application will be populated with information on several actual persons. Each particular person's individual data is an instance or item of the model.
# Property Values
Every instance of a model is grouping values for the properties defined for the model. Either value has to comply with the type and constraints as defined.
Example
In address book application there might be a record for a person named "John Doe". This record will have a value "Doe" for the property named last name which is a string of characters and is present, thus complying with the definition of this property.