DRAFT: Body and Teaser are now separate fields

This page is a DRAFT text for inclusion in http://drupal.org/node/224333 when complete. It will be unpublished once it has been reviewed and incorporated into the main handbook page.

Node Body and Teaser Changes

In Drupal 7 the node body becomes a field in core. This means the body field has changed from a specially handled node field to a field handled with the Field API.

This change results in the following:

• The body field is a field in core of type: Long text and summary (text_with_summary)
• The $node->teaser field no longer exists
• The $node->body field still exists but no longer represents the full content of the node, it is now a display-only field containing only the body field contents. Additionally, its data-type changes from text to array. Changes made to this field will not be rendered so it is recommended developers simply ignore it.
• The body field should be viewed and/or changed in $node->content['body'] using Field API best practice
• The default node data entry form has both Summary and Full text form fields
• The JavaScript splitter is no longer used on the node data entry form (If you want to recreate this behavior an example can be found in the Drupal 6 Node Comment module)
• The data entry form processing rules for the Summary and Full text form fields are:
  • If Summary text exists, it is used for the Teaser view
  • If no Summary text exists, a trimmed version of Full text is used
  • The length of the trimmed version is set in the node's display settings
  • The length of the Full text trimmed version can be manually set with <!--break-->

If your module changes user-facing behavior on node data entry forms familiarize yourself with the Field API widgets and formatters.

The following functions have been removed:

• node_prepare(). $node->body and $node->teaser no longer need special preparation.
• node_body_field(). The form element 'body' is now generated by Field API
• node_teaser() has been renamed to text_summary(). The function now runs htmlcorrector if applicable.