Models
The content of this page might not be fully up-to-date with Strapi 5 yet.
As Strapi is a headless Content Management System (CMS), creating a data structure for the content is one of the most important aspects of using the software. Models define a representation of the data structure.
There are 2 different types of models in Strapi:
- content-types, which can be collection types or single types, depending on how many entries they manage,
- and components that are data structures re-usable in multiple content-types.
If you are just starting out, it is convenient to generate some models with the Content-type Builder directly in the admin panel. The user interface takes over a lot of validation tasks and showcases all the options available to create the content's data structure. The generated model mappings can then be reviewed at the code level using this documentation.
Model creationβ
Content-types and components models are created and stored differently.
Content-typesβ
Content-types in Strapi can be created:
- with the Content-type Builder in the admin panel,
- or with Strapi's interactive CLI
strapi generate
command.
The content-types use the following files:
schema.json
for the model's schema definition. (generated automatically, when creating content-type with either method)lifecycles.js
for lifecycle hooks. This file must be created manually.
These models files are stored in ./src/api/[api-name]/content-types/[content-type-name]/
, and any JavaScript or JSON file found in these folders will be loaded as a content-type's model (see project structure).
In TypeScript-enabled projects, schema typings can be generated using the ts:generate-types
command.
Componentsβ
Component models can't be created with CLI tools. Use the Content-type Builder or create them manually.
Components models are stored in the ./src/components
folder. Every component has to be inside a subfolder, named after the category the component belongs to (see project structure).
Model schemaβ
The schema.json
file of a model consists of:
- settings, such as the kind of content-type the model represents or the table name in which the data should be stored,
- information, mostly used to display the model in the admin panel and access it through the REST and GraphQL APIs,
- attributes, which describe the data structure of the model,
- and options used to defined specific behaviors on the model.
Model settingsβ
General settings for the model can be configured with the following parameters:
Parameter | Type | Description |
---|---|---|
collectionName | String | Database table name in which the data should be stored |
kind Optional, only for content-types | String | Defines if the content-type is:
|
// ./src/api/[api-name]/content-types/restaurant/schema.json
{
"kind": "collectionType",
"collectionName": "Restaurants_v1",
}
Model informationβ
The info
key in the model's schema describes information used to display the model in the admin panel and access it through the Content API. It includes the following parameters:
Parameter | Type | Description |
---|---|---|
displayName | String | Default name to use in the admin panel |
singularName | String | Singular form of the content-type name. Used to generate the API routes and databases/tables collection. Should be kebab-case. |
pluralName | String | Plural form of the content-type name. Used to generate the API routes and databases/tables collection. Should be kebab-case. |
description | String | Description of the model |
"info": {
"displayName": "Restaurant",
"singularName": "restaurant",
"pluralName": "restaurants",
"description": ""
},
Model attributesβ
The data structure of a model consists of a list of attributes. Each attribute has a type
parameter, which describes its nature and defines the attribute as a simple piece of data or a more complex structure used by Strapi.
Many types of attributes are available:
- scalar types (e.g. strings, dates, numbers, booleans, etc.),
- Strapi-specific types, such as:
media
for files uploaded through the Media libraryrelation
to describe a relation between content-typescustomField
to describe custom fields and their specific keyscomponent
to define a component (i.e. a data structure usable in multiple content-types)dynamiczone
to define a dynamic zone (i.e. a flexible space based on a list of components)- and the
locale
andlocalizations
types, only used by the Internationalization (i18n) plugin
The type
parameter of an attribute should be one of the following values:
Type categories | Available types |
---|---|
String types |
|
Date types |
|
Number types |
|
Other generic types |
|
Special types unique to Strapi | |
Internationalization (i17n)-related types Can only be used if the i18n plugin is installed |
|
Validationsβ
Basic validations can be applied to attributes using the following parameters:
Parameter | Type | Description | Default |
---|---|---|---|
required | Boolean | If true , adds a required validator for this property | false |
max | Integer | Checks if the value is greater than or equal to the given maximum | - |
min | Integer | Checks if the value is less than or equal to the given minimum | - |
minLength | Integer | Minimum number of characters for a field input value | - |
maxLength | Integer | Maximum number of characters for a field input value | - |
private | Boolean | If true , the attribute will be removed from the server response.π‘ This is useful to hide sensitive data. | false |
configurable | Boolean | If false , the attribute isn't configurable from the Content-type Builder plugin. | true |
{
// ...
"attributes": {
"title": {
"type": "string",
"minLength": 3,
"maxLength": 99,
"unique": true
},
"description": {
"default": "My description",
"type": "text",
"required": true
},
"slug": {
"type": "uid",
"targetField": "title"
}
// ...
}
}
Database validations and settingsβ
These settings should be reserved to an advanced usage, as they might break some features. There are no plans to make these settings stable.
Database validations and settings are custom options passed directly onto the tableBuilder
Knex.js function during schema migrations. Database validations allow for an advanced degree of control for setting custom column settings. The following options are set in a column: {}
object per attribute:
Parameter | Type | Description | Default |
---|---|---|---|
name | string | Changes the name of the column in the database | - |
defaultTo | string | Sets the database defaultTo , typically used with notNullable | - |
notNullable | boolean | Sets the database notNullable , ensures that columns cannot be null | false |
unsigned | boolean | Only applies to number columns, removes the ability to go negative but doubles maximum length | false |
unique | boolean | Enforces database level unique, caution when using with draft & publish feature | false |
type | string | Changes the database type, if type has arguments, you should pass them in args | - |
args | array | Arguments passed into the Knex.js function that changes things like type | [] |
{
// ...
"attributes": {
"title": {
"type": "string",
"minLength": 3,
"maxLength": 99,
"unique": true,
"column": {
"unique": true // enforce database unique also
}
},
"description": {
"default": "My description",
"type": "text",
"required": true,
"column": {
"defaultTo": "My description", // set database level default
"notNullable": true // enforce required at database level, even for drafts
}
},
"rating": {
"type": "decimal",
"default": 0,
"column": {
"defaultTo": 0,
"type": "decimal", // using the native decimal type but allowing for custom precision
"args": [
6,1 // using custom precision and scale
]
}
}
// ...
}
}
uid
typeβ
The uid
type is used to automatically prefill the field value in the admin panel with a unique identifier (UID) (e.g. slugs for articles) based on 2 optional parameters:
targetField
(string): If used, the value of the field defined as a target is used to auto-generate the UID.options
(string): If used, the UID is generated based on a set of options passed to the underlyinguid
generator. The resultinguid
must match the following regular expression pattern:/^[A-Za-z0-9-_.~]*$
.
Relationsβ
Relations link content-types together. Relations are explicitly defined in the attributes of a model with type: 'relation'
and accept the following additional parameters:
Parameter | Description |
---|---|
relation | The type of relation among these values:
|
target | Accepts a string value as the name of the target content-type |
mappedBy and inversedBy Optional | In bidirectional relations, the owning side declares the inversedBy key while the inversed side declares the mappedBy key |
- One-to-one
- One-to-Many
- Many-to-One
- Many-to-Many
One-to-One relationships are useful when one entry can be linked to only one other entry.
They can be unidirectional or bidirectional. In unidirectional relationships, only one of the models can be queried with its linked item.
Unidirectional use case example:
- A blog article belongs to a category.
- Querying an article can retrieve its category,
- but querying a category won't retrieve the owned article.
// β¦
attributes: {
category: {
type: 'relation',
relation: 'oneToOne',
target: 'category',
},
},
// β¦
Bidirectional use case example:
- A blog article belongs to a category.
- Querying an article can retrieve its category,
- and querying a category also retrieves its owned article.
// β¦
attributes: {
category: {
type: 'relation',
relation: 'oneToOne',
target: 'category',
inversedBy: 'article',
},
},
// β¦
// β¦
attributes: {
article: {
type: 'relation',
relation: 'oneToOne',
target: 'article',
mappedBy: 'category',
},
},
// β¦
One-to-Many relationships are useful when:
- an entry from a content-type A is linked to many entries of another content-type B,
- while an entry from content-type B is linked to only one entry of content-type A.
One-to-many relationships are always bidirectional, and are usually defined with the corresponding Many-to-One relationship:
Details
Example:
A person can own many plants, but a plant is owned by only one person.
// β¦
attributes: {
owner: {
type: 'relation',
relation: 'manyToOne',
target: 'api::person.person',
inversedBy: 'plants',
},
},
// β¦
// β¦
attributes: {
plants: {
type: 'relation',
relation: 'oneToMany',
target: 'api::plant.plant',
mappedBy: 'owner',
},
},
// β¦
Many-to-One relationships are useful to link many entries to one entry.
They can be unidirectional or bidirectional. In unidirectional relationships, only one of the models can be queried with its linked item.
Unidirectional use case example:
A book can be written by many authors.
// β¦
attributes: {
author: {
type: 'relation',
relation: 'manyToOne',
target: 'author',
},
},
// β¦
Bidirectional use case example:
An article belongs to only one category but a category has many articles.
// β¦
attributes: {
author: {
type: 'relation',
relation: 'manyToOne',
target: 'category',
inversedBy: 'article',
},
},
// β¦
// β¦
attributes: {
books: {
type: 'relation',
relation: 'oneToMany',
target: 'article',
mappedBy: 'category',
},
},
// β¦
Many-to-Many relationships are useful when:
- an entry from content-type A is linked to many entries of content-type B,
- and an entry from content-type B is also linked to many entries from content-type A.
Many-to-many relationships can be unidirectional or bidirectional. In unidirectional relationships, only one of the models can be queried with its linked item.
Unidirectional use case example:
// β¦
attributes: {
categories: {
type: 'relation',
relation: 'manyToMany',
target: 'category',
},
},
// β¦
Bidirectional use case example:
An article can have many tags and a tag can be assigned to many articles.
// β¦
attributes: {
tags: {
type: 'relation',
relation: 'manyToMany',
target: 'tag',
inversedBy: 'articles',
},
},
// β¦
// β¦
attributes: {
articles: {
type: 'relation',
relation: 'manyToMany',
target: 'article',
mappedBy: 'tag',
},
},
// β¦
Custom fieldsβ
Custom fields extend Strapiβs capabilities by adding new types of fields to content-types. Custom fields are explicitly defined in the attributes of a model with type: customField
.
Custom fields' attributes also accept:
Custom fields' attributes also show the following specificities:
- a
customField
attribute whose value acts as a unique identifier to indicate which registered custom field should be used. Its value follows:- either the
plugin::plugin-name.field-name
format if a plugin created the custom field - or the
global::field-name
format for a custom field specific to the current Strapi application
- either the
- and additional parameters depending on what has been defined when registering the custom field (see custom fields documentation).
{
// β¦
"attributes": {
"attributeName": { // attributeName would be replaced by the actual attribute name
"type": "customField",
"customField": "plugin::color-picker.color",
"options": {
"format": "hex"
}
}
}
// β¦
}
Componentsβ
Component fields create a relation between a content-type and a component structure. Components are explicitly defined in the attributes of a model with type: 'component'
and accept the following additional parameters:
Parameter | Type | Description |
---|---|---|
repeatable | Boolean | Could be true or false depending on whether the component is repeatable or not |
component | String | Define the corresponding component, following this format:<category>.<componentName> |
{
"attributes": {
"openinghours": {
"type": "component",
"repeatable": true,
"component": "restaurant.openinghours"
}
}
}
Dynamic zonesβ
Dynamic zones create a flexible space in which to compose content, based on a mixed list of components.
Dynamic zones are explicitly defined in the attributes of a model with type: 'dynamiczone'
. They also accept a components
array, where each component should be named following this format: <category>.<componentName>
.
{
"attributes": {
"body": {
"type": "dynamiczone",
"components": ["article.slider", "article.content"]
}
}
}
Model optionsβ
The options
key is used to define specific behaviors and accepts the following parameter:
Parameter | Type | Description |
---|---|---|
privateAttributes | Array of strings | Allows treating a set of attributes as private, even if they're not actually defined as attributes in the model. It could be used to remove them from API responses timestamps. The privateAttributes defined in the model are merged with the privateAttributes defined in the global Strapi configuration. |
draftAndPublish | Boolean | Enables the draft and publish feature. Default value: true (false if the content-type is created from the interactive CLI). |
{
"options": {
"privateAttributes": ["id", "createdAt"],
"draftAndPublish": true
}
}
Lifecycle hooksβ
Lifecycle hooks are functions that get triggered when Strapi queries are called. They are triggered automatically when managing content through the administration panel or when developing custom code using queries
Β·
Lifecycle hooks can be customized declaratively or programmatically.
Lifecycles hooks are not triggered when using directly the knex library instead of Strapi functions.
The Document Service API triggers various database lifecycle hooks based on which method is called. For a complete reference, see Document Service API: Lifecycle hooks. Bulk actions lifecycles (createMany
, updateMany
, deleteMany
) will never be triggered by a Document Service API method. Document Service middlewares can be implemented too.
Available lifecycle eventsβ
The following lifecycle events are available:
beforeCreate
beforeCreateMany
afterCreate
afterCreateMany
beforeUpdate
beforeUpdateMany
afterUpdate
afterUpdateMany
beforeDelete
beforeDeleteMany
afterDelete
afterDeleteMany
beforeCount
afterCount
beforeFindOne
afterFindOne
beforeFindMany
afterFindMany
Hook event
objectβ
Lifecycle hooks are functions that take an event
parameter, an object with the following keys:
Key | Type | Description |
---|---|---|
action | String | Lifecycle event that has been triggered (see list) |
model | Array of strings (uid) | An array of uids of the content-types whose events will be listened to. If this argument is not supplied, events are listened on all content-types. |
params | Object | Accepts the following parameters:
|
result | Object | Optional, only available with afterXXX eventsContains the result of the action. |
state | Object | Query state, can be used to share state between beforeXXX and afterXXX events of a query. |
Declarative and programmatic usageβ
To configure a content-type lifecycle hook, create a lifecycles.js
file in the ./src/api/[api-name]/content-types/[content-type-name]/
folder.
Each event listener is called sequentially. They can be synchronous or asynchronous.
- JavaScript
- TypeScript
module.exports = {
beforeCreate(event) {
const { data, where, select, populate } = event.params;
// let's do a 20% discount everytime
event.params.data.price = event.params.data.price * 0.8;
},
afterCreate(event) {
const { result, params } = event;
// do something to the result;
},
};
export default {
beforeCreate(event) {
const { data, where, select, populate } = event.params;
// let's do a 20% discount everytime
event.params.data.price = event.params.data.price * 0.8;
},
afterCreate(event) {
const { result, params } = event;
// do something to the result;
},
};
Using the database layer API, it's also possible to register a subscriber and listen to events programmatically:
module.exports = {
async bootstrap({ strapi }) {
// registering a subscriber
strapi.db.lifecycles.subscribe({
models: [], // optional;
beforeCreate(event) {
const { data, where, select, populate } = event.params;
event.state = 'doStuffAfterWards';
},
afterCreate(event) {
if (event.state === 'doStuffAfterWards') {
}
const { result, params } = event;
// do something to the result
},
});
// generic subscribe for generic handling
strapi.db.lifecycles.subscribe((event) => {
if (event.action === 'beforeCreate') {
// do something
}
});
}
}