search-it-extension-points

name	search-it-extension-points
description	Search It extension points – SEARCH_IT_INDEX_ARTICLE, SEARCH_IT_PLAINTEXT, SEARCH_IT_SEARCH_EXECUTED. Use when the user wants to hook into search indexing, customise plaintext conversion, exclude articles from the index, track search analytics, or modify search behaviour via rex_extension.

Search It – Extension Points

Search It provides three extension points for customising indexing and search behaviour. Register them in your project addon's boot.php.

SEARCH_IT_INDEX_ARTICLE

Fires before an article is indexed. Return false to exclude the article from the index.

Parameters:

Param	Type	Description
article	rex_article	The article about to be indexed

Exclude articles by meta field

rex_extension::register('SEARCH_IT_INDEX_ARTICLE', function(rex_extension_point $ep) {
    $article = $ep->getParam('article');

    // Skip articles with noindex meta field
    if ($article->getValue('art_noindex') == 1) {
        return false;
    }
});

Exclude offline categories

rex_extension::register('SEARCH_IT_INDEX_ARTICLE', function(rex_extension_point $ep) {
    $article = $ep->getParam('article');
    $category = $article->getCategory();

    if ($category && !$category->isOnline()) {
        return false;
    }
});

Exclude specific templates

rex_extension::register('SEARCH_IT_INDEX_ARTICLE', function(rex_extension_point $ep) {
    $article = $ep->getParam('article');
    $excludedTemplates = [5, 8]; // template IDs not to index

    if (in_array($article->getTemplateId(), $excludedTemplates)) {
        return false;
    }
});

SEARCH_IT_PLAINTEXT

Fires during plaintext conversion, after the article HTML has been fetched but before the standard plaintext processing. Allows you to modify or replace the text that gets indexed.

Subject: string – the raw HTML/text content

Remove specific elements before indexing

rex_extension::register('SEARCH_IT_PLAINTEXT', function(rex_extension_point $ep) {
    $text = $ep->getSubject();

    // Remove elements that should not appear in search results
    $text = preg_replace('/<nav[^>]*>.*?<\/nav>/si', '', $text);
    $text = preg_replace('/<div class="no-search">.*?<\/div>/si', '', $text);

    return $text;
});

Return array to control processing pipeline

rex_extension::register('SEARCH_IT_PLAINTEXT', function(rex_extension_point $ep) {
    $text = $ep->getSubject();

    // Custom processing
    $text = strip_tags($text, '<p><br>');
    $text = html_entity_decode($text, ENT_QUOTES, 'UTF-8');

    return [
        'text' => $text,
        'process' => false  // false = skip built-in plaintext conversion
                            // true  = run built-in conversion after this hook
    ];
});

Add extra content to the index

rex_extension::register('SEARCH_IT_PLAINTEXT', function(rex_extension_point $ep) {
    $text = $ep->getSubject();
    $params = $ep->getParams();

    // Append YForm data related to this article
    $articleId = $params['article_id'] ?? null;
    if ($articleId) {
        $extra = rex_sql::factory()
            ->setQuery('SELECT description FROM ' . rex::getTable('my_data') . ' WHERE article_id = ?', [$articleId]);

        if ($extra->getRows() > 0) {
            $text .= ' ' . $extra->getValue('description');
        }
    }

    return $text;
});

SEARCH_IT_SEARCH_EXECUTED

Fires after every search query is executed. Useful for logging, analytics, or modifying results.

Subject: array – the full result array (same structure as $search->search() returns)

Log searches to a custom table

rex_extension::register('SEARCH_IT_SEARCH_EXECUTED', function(rex_extension_point $ep) {
    $result = $ep->getSubject();

    rex_sql::factory()
        ->setTable(rex::getTable('search_log'))
        ->setValue('searchterm', $result['searchterm'])
        ->setValue('results_count', $result['count'])
        ->setValue('search_time', $result['time'])
        ->setValue('createdate', date('Y-m-d H:i:s'))
        ->insert();
});

Track zero-result searches

rex_extension::register('SEARCH_IT_SEARCH_EXECUTED', function(rex_extension_point $ep) {
    $result = $ep->getSubject();

    if ($result['count'] === 0) {
        rex_logger::factory()->log('search_it', 'Zero results for: ' . $result['searchterm']);
    }
});

Modify results (filter, reorder)

rex_extension::register('SEARCH_IT_SEARCH_EXECUTED', function(rex_extension_point $ep) {
    $result = $ep->getSubject();

    // Remove file hits from results
    $result['hits'] = array_filter($result['hits'], function($hit) {
        return $hit['type'] !== 'file';
    });

    $result['hits'] = array_values($result['hits']); // reindex
    return $result;
});

Registration location

Always register these extension points in your project addon's boot.php (or a custom addon's boot.php), never in a module or template:

// redaxo/src/addons/project/boot.php

rex_extension::register('SEARCH_IT_INDEX_ARTICLE', function(rex_extension_point $ep) {
    // ...
});

Common pitfalls

• Registering extension points in a module output – they run only when that article is displayed, not during indexing. Always use boot.php.
• Returning null instead of false in SEARCH_IT_INDEX_ARTICLE – null does not exclude the article. You must explicitly return false.
• Using heavy database queries in SEARCH_IT_PLAINTEXT – this runs for every article during reindex. Keep it lightweight or cache results.
• Modifying $result['hits'] in SEARCH_IT_SEARCH_EXECUTED without updating $result['count'] – pagination breaks if count no longer matches actual hits.
• Forgetting return in SEARCH_IT_SEARCH_EXECUTED when modifying results – without returning the modified array, changes are lost.