Documentation AEM as a Cloud Service User Guide

Content Fragment Models

Last update: February 24, 2025

CREATED FOR:

User
Admin
Developer

Content Fragment Models in AEM define the structure of content for your content fragments, serving as a foundation of your headless content.

To use Content Fragment Models you:

Enable Content Fragment Model functionality for your instance
Create, and configure, your Content Fragment Models
Enable your Content Fragment Models for use when creating Content Fragments
Allow your Content Fragment Models on the required Assets folders by configuring Policies.

NOTE

Content Fragments are a Sites feature, but are stored as Assets.

Content Fragments and Content Fragment Models are now primarily managed with the Content Fragments console, though Content Fragments can still be managed from the Assets console, and Content Fragment Models from the Tools console. This section covers management from the Assets and Tools consoles.

Creating a Content Fragment Model

Navigate to Tools, General, then open Content Fragment Models.
Navigate to the folder appropriate to your configuration, or sub-configuration.
Use Create to open the wizard.

CAUTION
If the use of content fragment models have not been enabled, the Create option will not be available.
Specify the Model Title.
You can also define various properties; for example, add Tags, a Description, and select Enable model to enable the model if necessary.

NOTE
For details about the Default Preview URL Pattern see Content Fragment Model - Properties.
Use Create to save the empty model. A message will indicate the success of the action, you can select Open to immediately edit the model, or Done to return to the console.

Defining your Content Fragment Model

The content fragment model effectively defines the structure of the resulting content fragments using a selection of Data Types. Using the model editor you can add instances of the data types, then configure them to create the required fields:

CAUTION

Editing an existing content fragment model can impact dependent fragments.

Navigate to Tools, General, then open Content Fragment Models.
Navigate to the folder holding your content fragment model.
Open the required model for Edit; use either the quick action, or select the model and then the action from the toolbar.

Once open the model editor shows:
- left: fields already defined
- right: Data Types available for creating fields (and Properties for use once fields have been created)
NOTE
When a field as Required, the Label indicated in the left pane is marked with an asterix (*).

properties

To Add a Field
- Drag a required data type to the required location for a field:
- Once a field has been added to the model, the right panel will show the Properties that can be defined for that particular data type. Here you can define what is required for that field.
  - Many properties are self-explanatory, for additional details see Properties.
  - Typing a Field Label will auto-complete the Property Name - if empty, and it can be manually updated afterwards.
    
    CAUTION
    When manually updating the property Property Name for a data type, note that names must contain only A-Z, a-z, 0-9 and underscore “_” as special character.
    If models created in earlier versions of AEM contain illegal characters, remove or update those characters.
  For example:
To Remove a Field

Select the required field, then select the trash-can icon. You are asked to confirm the action.
Add all required fields, and define the related properties, as required. For example:
Select Save to persist the definition.

Data Types

A selection of data types is available for defining your model:

Single line text
- Add a field for a single line of text; the maximum length can be defined
- The field can be configured to allow fragment authors to create new instances of the field
Multi line text
- A text area that can be Rich Text, Plain Text, or Markdown
- The field can be configured to allow fragment authors to create new instances of the field
NOTE
Whether the text area is Rich Text, Plain Text, or Markdown, is defined in the model by the property Default Type.
This format cannot be changed from the Content Fragment editor, but only from the Model.
Number
- Add a numerical field
- The field can be configured to allow fragment authors to create new instances of the field
Boolean
- Add a boolean checkbox
Date and time
- Add a date and/or time field
Enumeration
- Add a set of Checkbox, Radio Button, or Dropdown fields
  - You can specify the options available to the fragment author
Tags
- Allows fragment authors to access and select areas of tags
Content Reference
- References other content, of any type; can be used to create nested content
- If an image is referenced, you can opt to show a thumbnail
- The field can be configured to allow fragment authors to create new instances of the field
Fragment Reference
- References other Content Fragments; can be used to create nested content
- The field can be configured to allow fragment authors to:
  - Edit the referenced fragment directly
  - Create a new Content Fragment, based on the appropriate model
  - Create new instances of the field
JSON Object
- Allows the Content Fragment author to enter JSON syntax into the corresponding elements of a fragment.
  - To allow AEM to store direct JSON that you have copy/pasted from another service.
  - The JSON is passed through, and output as JSON in GraphQL.
  - Includes JSON syntax-highlighting, auto-complete, and error-highlighting in the Content Fragment editor.
Tab Placeholder
- Allows the introduction of tabs for use when editing the Content Fragment content.
  - These are shown as dividers in the model editor, separating sections of the list of content data types. Each instance represents the start of a new tab.
  - In the fragment editor each instance appears as a tab.
  NOTE
  This data type is purely used for formatting, it is ignored by the AEM GraphQL schema.

Properties

Many properties are self-explanatory, for certain properties additional details are below:

Property Name

When manually updating this property for a data type, note that names must contain only A-Z, a-z, 0-9 and underscore “_” as special character.

CAUTION
If models created in earlier versions of AEM contain illegal characters, remove or update those characters.
Render As
The various options for realizing/rendering the field in a fragment. Often this property lets you define whether the author sees a single instance of the field, or is allowed to create multiple instances. When Multiple Field is used you can define the minimum and maximum number of items - see Validation for further details.
Field Label
Entering a Field Label will autogenerate a Property Name, which can then be manually updated if necessary.
Validation
Basic validation is available by mechanisms such as the Required property. Some data types have addition validation fields. See Validation for further details.
For the data type Multi line text it is possible to define the Default Type as either:
- Rich Text
- Markdown
- Plain Text
If not specified, the default value Rich Text is used for this field.

Changing the Default Type in a content fragment model will only take effect on an existing, related, content fragment after that fragment is opened in the editor and saved.
Unique
Content (for the specific field) must be unique across all content fragments created from the current model.

This is used to ensure that content authors cannot repeat content already added in another fragment of the same model.

For example, a Single line text field called Country in the Content Fragment Model cannot have the value Japan in two dependent Content Fragments. A warning is issued when the second instance is attempted.

NOTE
Uniqueness is ensured per language root.

NOTE
Variations can have the same unique value as variations of the same fragment, but not the same value as used in any variation of other fragments.

CAUTION
If you want to use MSM (which creates copies of Content Fragments), then any Unique constraints should be removed from any Data Types used in the respective Content Fragment Models.
See Content Reference for more details about that specific data type and its properties.
See Fragment Reference (Nested Fragments) for more details about that specific data type and its properties.
Translatable

Checking the Translatable checkbox on a field in the Content Fragment Model editor will:
- Ensure the field’s property name is added to the translation configuration, context /content/dam/<sites-configuration>, if not already present.
- For GraphQL: set a <translatable> property on the Content Fragment field to yes, to allow GraphQL query filter for JSON output with only translatable content.

Validation

Various data types now include the possibility to define validation requirements for when content is entered in the resulting fragment:

Single line text
- Compare against a predefined regex.
Number
- Check for specific values.
Content Reference
- Test for specific types of content.
- Only assets of specified file size or smaller can be referenced.
- Only images within a predefined range of width and/or height (in pixels) can be referenced.
Fragment Reference
- Test for a specific content fragment model.
Min Number of Items / Max Number of Items

Fields that have been defined as a Multiple Field (set with Render As) have the options:
- Min Number of Items
- Max Number of Items
These are validated:
- The max value is validated in the original Content Fragment Editor.
- Both are validated in the Content Fragment Editor.

Using References to form Nested Content

Content Fragments can form nested content, using either of the following data types:

Content Reference
- Provides a simple reference to other content; of any type.
- Can be configured for a one or multiple references (in the resulting fragment).
Fragment Reference (Nested Fragments)
- References other fragments, dependent on the specific models specified.
- Lets you include/retrieve structured data.
  
  NOTE
  This method is of particular interest in conjunction with Headless Content Delivery using Content Fragments with GraphQL.
  
  * Can be configured for one or multiple references (in the resulting fragment)..

NOTE

AEM has a recurrence protection for:

Content References
This prevents the user from adding a reference to the current fragment. This may lead to an empty Fragment Reference picker dialog.
Fragment References in GraphQL
If you create a deep query that returns multiple Content Fragments referenced by each other, it will return null at first occurrence.

Content Reference

The Content Reference lets you render content from another source; for example, image or content fragment.

In addition to standard properties you can specify:

The Root Path for any referenced content
The content types that can be referenced
Limitations for file sizes
If an image is referenced:
- Show Thumbnail
- Image restraints of height and width

Content Reference

Fragment Reference (Nested Fragments)

The Fragment Reference references one, or more, content fragments. This feature of particular interest when retrieving content for use in your app, as it lets you retrieve structured data with multiple layers.

For example:

A model defining details for an employee; these include:
- A reference to the model that defines the employer (company)

type EmployeeModel {
    name: String
    firstName: String
    company: CompanyModel
}

type CompanyModel {
    name: String
    street: String
    city: String
}

NOTE

This is of particular interest in conjunction with Headless Content Delivery using Content Fragments with GraphQL.

In addition to standard properties you can define:

Render As:
- multifield - the fragment author can create multiple, individual, references
- fragmentreference - allows the fragment author to select a single reference to a fragment
Model Type
Multiple models can be selected. When authoring the Content Fragment any referenced fragments must have been created using these models.
Root Path
This specifies a root path for any fragments referenced.
Allow Fragment Creation

This allows the fragment author to create a fragment based on the appropriate model.
- fragmentreferencecomposite - allows the fragment author to build a composite, by selecting multiple fragments

NOTE

A recurrence protection mechanism is in place. It prohibits the user from selecting the current Content Fragment in the Fragment Reference. This may lead to an empty Fragment Reference picker dialog.

There is also a recurrence protection for Fragment References in GraphQL. If you create a deep query across two Content Fragments that reference each other, it will return null.

Content Fragment Model - Properties

You can edit the Properties of a Content Fragment Model:

Basic
- Model Title
- Tags
- Description
- Upload Image
- Default Preview URL Pattern
  
  NOTE
  This is only used by the new Content Fragment Editor. See Content Fragment Models for further information.

Enabling or Disabling a Content Fragment Model

For full control over the use of your Content Fragment Models they have a status that you can set.

Enabling a Content Fragment Model

When a model is created it must be enabled so that it:

Is available for selection when creating a Content Fragment.
Can be referenced from within a Content Fragment Model.
Is available to GraphQL; so the schema is generated.

To enable a Model that is flagged as either:

Draft : mew (never enabled).
Disabled : has been specifically disabled.

You use the Enable option from either:

The top toolbar, when the required Model is selected.
The corresponding Quick Action (mouse-over the required Model).

Enable a Draft or Disabled Model

Disabling a Content Fragment Model

A model can also be disabled so that:

The model is no longer available as a basis for creating new Content Fragments.
However:
- The GraphQL schema keeps being generated and is still queryable (to avoid impacting JSON API).
- Any Content Fragments based of the model can still be queried and returned from the GraphQL endpoint.
The model cannot be referenced anymore, but existing references are kept untouched, and can still be queried and returned from the GraphQL endpoint.

To disable a Model that is flagged as Enabled you use the Disable option from either:

The top toolbar, when the required Model is selected.
The corresponding Quick Action (mouse-over the required Model).

Disable an Enabled Model

Allowing Content Fragment Models on your Assets Folder

To implement content governance, you can configure Policies on Assets folder to control which Content Fragment Models are allowed for Fragment creation in that folder.

NOTE

The mechanism is similar to allowing page templates for a page, and its children, in advanced properties of a page.

To configure the Policies for Allowed Content Fragment Models:

Navigate and open Properties for the required Assets folder.
Open the Policies tab, where you can configure:
- Inherited from <folder>
  
  Policies are automatically inherited when creating new child folders; the policy can be reconfigured (and the inheritance broken) if sub-folders need to allow models different to the parent folder.
- Allowed Content Fragment Models by Path
  
  Multiple models can be allowed.
- Allowed Content Fragment Models by Tag
  
  Multiple models can be allowed.
Save any changes.

The Content Fragment Models allowed for a folder are resolved as follows:

The Policies for Allowed Content Fragment Models.
If empty, then try to determine the policy using the inheritance rules.
If the inheritance chain does not deliver a result, then look at the Cloud Services configuration for that folder (also first directly and then via inheritance).
If none of the above deliver any results, then there are no allowed models for that folder.

Deleting a Content Fragment Model

CAUTION

Deleting a content fragment model can impact dependent fragments.

To delete a content fragment model:

Navigate to Tools, General, then open Content Fragment Models.
Navigate to the folder holding your content fragment model.
Select your model, followed by Delete from the toolbar.

NOTE
If the model is referenced, a warning is given. Take action appropriately.

Publishing a Content Fragment Model

Content fragment models need to be published when/before any dependent content fragments are published.

To publish a content fragment model:

Navigate to Tools, General, then open Content Fragment Models.
Navigate to the folder holding your content fragment model.
Select your model, followed by Publish from the toolbar.
The published status is indicated in the console.

NOTE
If you publish a content fragment for which the model has not yet been published, a selection list indicates this and the model is published with the fragment.

Unpublishing a Content Fragment Model

Content fragment models can be unpublished if they are not referenced by any fragments.

To unpublish a content fragment model:

Navigate to Tools, General, then open Content Fragment Models.
Navigate to the folder holding your content fragment model.
Select your model, followed by Unpublish from the toolbar.
The published status is indicated in the console.

If you try to unpublish a model that is currently used by one or more fragments, then an error warning will inform you of this:

Content Fragment Model error message when unpublishing a model that is in use

The message will suggest that you check the References panel to investigate further:

Content Fragment Model in References

Locked (Published) Content Fragment Models

This feature provides governance for Content Fragment Models that have been published.

The Challenge

Content Fragment Models determine the schema for GraphQL queries in AEM.
- AEM GraphQL schemas are created as soon as a Content Fragment Model is created, and they can exist on both author and publish environments.
- Schemas on publish are the most critical as they provide the foundation for live delivery of Content Fragment content in JSON format.
Problems can occur when Content Fragment Models are modified, or in other words edited. This means that the schema changes, which in turn may affect existing GraphQL queries.
Adding new fields to a Content Fragment Model should (typically) not have any detrimental effects. However, modifying existing data fields (for example, their name) or deleting field definitions, will break existing GraphQL queries when they are requesting these fields.

The Requirements

To make users aware of the risks when editing models that are already used for live content delivery - in other words, models that have been published.
Also, to avoid unintended changes.

Either of these might break queries if the modified models are re-published.

The Solution

To address these issues, Content Fragment Models are locked into a READ-ONLY mode on author - as soon as they have been published. This is indicated by Locked:

Card of locked Content Fragment Model

When the model is Locked (in READ-ONLY mode), you can see the contents and structure of models but you cannot edit them.

You can manage Locked models from either the console, or the model editor:

Console

From the console, you can manage the READ-ONLY mode with the Unlock and Lock actions in the toolbar:
- You can Unlock a model to enable edits.
  
  If you select Unlock, a warning is shown, and you must confirm the Unlock action:
  
  You can then open the model for editing.
- You can also Lock the model afterwards.
- Re-publishing the model will immediately put it back into Locked (READ-ONLY) mode.
Model Editor
- When you open a model that is locked, you are warned, and presented with three actions: Cancel, View Read Only, Edit:
- If you select View Read Only you can see the content and structure of the model:
- If you select Edit you can edit and save your updates:
  
  NOTE
  There may still a warning at the top, but that is when the model is already in use by existing Content Fragments.
- Cancel will return you to the console.

recommendation-more-help