Lesson 4. Schemas
Example 4.1. Implicit type specification
TYPE @cat
{ # implicitly set `object`
"id" : 123, # implicitly set `integer`
"name" : "Tom", # implicitly set `string`
"isLovely" : true , # implicitly set `boolean`
"weight" : 7.1, # implicitly set `float`
"friendIds": [ # implicitly set `array` of `integer`s [0–∞]
232
],
"address": { # implicitly set `object`
"country": "USA", # implicitly set `string`
"city" : "New York", # implicitly set `string`
"line 1" : "Cats street, 23", # implicitly set `string`
"zipcode": 123456 # implicitly set `integer`
}
}
Note. The hash mark # is followed by comments.
In this example, we demonstrate data types that are implicitly but clearly defined from an example
of valid data. These are the types: object, array, integer, float, string, boolean.
These types, if desired, can be specified explicitly using the rule type as follows:
TYPE @cat
{ // {type: "object"} # type is optional
"id": 123 // {type: "integer"} # type is optional
}
info
For more information about this topic, see:
Example 4.2. Other built-in types
TYPE @cat
{
"id" : 123,
"name" : "Tom",
"birthday" : "2006-01-02", // {type: "date"}
"registered": "2021-01-02T07:23:12+03:00", // {type: "datetime"}
"email" : "tom@cats.com", // {type: "email" }
"website" : "http://tom.cats.com", // {type: "uri" }
"salary" : 13.23, // {type: "decimal", precision: 2}
"color" : "white", // {type: "enum", enum: ["white", "black"]}
"uuid" : "550e8400-e29b-41d4-a716-446655440000" // {type: "uuid"}
}
JSight has several built-in types that must always be specified explicitly in the rule type. These
types are depicted in the example to the left. These types include:
decimal,date,datetime,email,uri,enum,uuid.
Types decimal and enum always require additional rules (type decimal requires the rule
precision, and type enum requires the rule enum).
Therefore, for these types, the rule type can be omitted, especially as it has no effect on
perception:
TYPE @cat
{
"salary": 13.23, // {precision: 2} # type is `decimal`
"color" : "white" // {enum: ["white", "black"]} # type is `enum`
}
info
For more information about this topic, see:
Example 4.3. User types
TYPE @cat
{
"id" : @catId,
"name" : "Tom",
"friends": [@cat]
}
TYPE @catId
"CAT-123" // {regex: "^CAT-\\d+$"}
In the given example, we can see how user types can be specified in schemas:
In the type @cat the user type @catId is specified for the field "id" (line 3).
In the type @cat the elements of array "friends" are assigned to the type @cat (line 5).
info
For more information about this topic, see:
Example 4.4. User types – expressed with the example
TYPE @cat
{
"id" : "CAT-23454", // {type: "@catId"}
"name" : "Tom",
"friends": [@cat]
}
TYPE @catId
"CAT-123" // {regex: "^CAT-\\d+$"}
In the given example, the same type of data is provided, but we specify an example value for the
field "id" in the type @cat, while the reference to the user data type was indicated in the
annotation in the rule type (string 3).
This is convenient to do in the case of scalar types. This cannot be done with arrays and objects.
info
Example 4.5. Reference to multiple user types
TYPE @cat
{
"id" : 123,
"name": "Tom",
"closestFriend": @cat | @dog
}
In the given example, the value of the property "closestFriend" must be of type @cat or @dog.
info
For more information about this topic, see:
A table of all built-in types and rules
The table below shows all types as well as all of the rules that can be applied to them.