README

A set of helper methods to facilitate searching elemental pages

Requirements

• Silverstripe CMS ^6
• Silverstripe Elemental ^6

Installation

composer require dnadesign/silverstripe-elemental-textcontentforsearch

The following YAML config will enable elements on every Page object, replacing the standard Content rich text field and add access to the getTextContentForSearch method

mysite/_config/elements.yml

Page:
  extensions:
    - DNADesign\Elemental\Extensions\ElementalPageExtension
    - DNADesign\Elemental\Extensions\ElementalPageTextSearchExtension

Usage

SolrIndex

We recommend you use the TextContentForSearch instead of the ElementForSearch method in your index. This avoids rendering the entire elemental area, which would include images and complex elements, which consumes a lot of resources during render and don't end up in the index anyway.

class ElementalSolrIndex extends SolrIndex
{
    public function init()
    {
        $this->addClass(Page::class);
        $this->addAllFulltextFields();
        /** @see ElementalPageTextContentSearchExtension::getTextContentForSearch */
        $this->addFulltextField('TextContentForSearch');
    }
}

Configuration

By default, the process will extract any Varchar, Text and HTMLText field from the elements. If you require any other type of fields to be considered for indexing, you ca update the config as follow:

Page:
  db_text_field_types:
    - CustomFieldType

You can exclude fields per element. By default, the Style and ExtraClass fields are excluded as they wouldn't bring any value to the search. To exclude a field, you can either set the config on the class or in the yaml:

DNADesign\Elemental\Models\ElementContent:
  exclude_fields_from_search:
    - Title

or

class MyCustomElement extends BaseElement
{
    private static $exclude_fields_from_search = [
        'Title'
    ];
}

In reverse, you can add a field in the same way.

DNADesign\Elemental\Models\ElementContent:
  include_fields_in_search:
    - AlternativeText

or

class MyCustomElement extends BaseElement
{
    private static $include_fields_in_search = [
        'AlternativeText'
    ];
}

In addition, you can implement a updateTextFieldsForSearch method on any element or extension to update the list of fields to include in search.

Some elements may have complex relationships that are not explored by the default process. You can implement addTextContentForSearch to concatenate extra content to the final string.

class MyCustomElement extends BaseElement
{
    private static $has_many = [
        'AccordionItems' => AccordionItem::class
    ];

    public function addTextContentForSearch()
    {
        $titles = [];
        foreach($this->AccordionItems() as $item) {
            $titles[] = $item->Title;
        }
        return implode(' ', $titles);
    }
}

Ultimately, you can manipulate the final string to be return by implementing updateTextContentForSearch method.

class MyCustomElement extends BaseElement
{
    public function updateTextContentForSearch($string)
    {
        // Remove any occurrence of forbiddenWord in the final string
        return = str_replace('forbiddenWord', '', $string);
    }
}

By default, every element is considered for indexing. If you wish to exclude an element from the search, you can use the config as follow:

DNADesign\Elemental\Models\ElementContent:
  exclude_content_from_search: true

or

class MyCustomElement extends BaseElement
{
    private static $exclude_content_from_search = true;
}

Debugging

As it is difficult to see what is being included in the solr index, the module exposes a read only text field with the content generated by the getTextContentForSearch method. You can hide this field by amending the config:

DNADesign\Elemental\Extensions\ElementalPageTextSearchExtension:
  show_text_content_search_in_cms: false

Notes:

• The introspective process will explore any VirtualElement and extract the parent element text fields

To do:

• Add an option to store the search string in the DB on save?