Development of the 1C-Bitrix FAQ module

FAQ Module Development for 1C-Bitrix

FAQ on Bitrix is most commonly built with an infoblock containing two fields — question and answer — and the bitrix:catalog.list component in a simplified version. Problems begin when there are more than 50 questions: no full-text search across answers, no section grouping with accordion, no analytics on which questions are viewed most often, no way to ask a question directly from the FAQ page. A specialized module solves each of these problems.

Data Model

Module vendor.faq:

• b_vendor_faq_category — categories: id, name, slug, sort, is_active, questions_count (denormalized counter)
• b_vendor_faq_question — questions: id, category_id, question, answer (HTML), slug, sort, views, is_active, is_featured, created_at, updated_at
• b_vendor_faq_vote — "helped/didn't help" votes: id, question_id, user_id, ip, value, created_at
• b_vendor_faq_search_log — search log: id, query, results_count, created_at

Full-Text Search

Bitrix's standard search via bitrix:search.title indexes pages, not records. FAQ needs instant search across questions and answers without page reload:

// AJAX search endpoint
public function search(string $query): array
{
    $query = mb_strtolower(trim($query));

    // Log the query for analytics
    SearchLogTable::add(['QUERY' => $query]);

    $result = QuestionTable::getList([
        'select' => ['ID', 'QUESTION', 'ANSWER', 'CATEGORY_ID'],
        'filter' => [
            'IS_ACTIVE' => 'Y',
            [
                'LOGIC' => 'OR',
                '%QUESTION' => $query,
                '%ANSWER'   => $query,
            ],
        ],
        'limit' => 10,
        'order' => ['IS_FEATURED' => 'DESC', 'VIEWS' => 'DESC'],
    ])->fetchAll();

    // Highlight matches in results
    return array_map(fn($q) => $this->highlight($q, $query), $result);
}

For large databases (1000+ questions), %LIKE% search in PostgreSQL is replaced with a full-text index (tsvector). The fts_vector column is updated by a trigger when a question or answer is changed.

View Counter and Rating

When a question is opened, the counter is incremented via AJAX:

QuestionTable::update($questionId, ['VIEWS' => new \Bitrix\Main\DB\SqlExpression('VIEWS + 1')]);

After reading, the user can rate the answer: "This helped" / "Didn't help". Vote statistics are used for two purposes: sorting (helpful questions rank higher) and identifying problematic answers (many negatives — needs rewriting).

Accordion Structure

The vendor:faq.list component renders questions grouped by category. The accordion is implemented in CSS (:details/:summary) without JavaScript — works without JS and indexes cleanly. When the GET parameter ?q=slug-of-question is present, the relevant section is expanded automatically via the open attribute.

"Ask a Question" Form

If the user does not find an answer in the search, the form invites them to ask a question. The question is saved to b_vendor_feedback_submission (integration with the feedback module) or directly to b_vendor_faq_question with a draft status for later publication. The administrator can publish it in the FAQ if the question is of broad interest.

Schema.org FAQPage

Microdata is important for search visibility:

<div itemscope itemtype="https://schema.org/FAQPage">
  <div itemscope itemprop="mainEntity" itemtype="https://schema.org/Question">
    <h3 itemprop="name">How do I process a return?</h3>
    <div itemscope itemprop="acceptedAnswer" itemtype="https://schema.org/Answer">
      <div itemprop="text">To process a return...</div>
    </div>
  </div>
</div>

The component automatically generates microdata for all published questions on the page.

Analytics

A report is built from b_vendor_faq_search_log:

• Top search queries with no results — what users look for but don't find
• Popular questions by views over a period
• Questions with a high "didn't help" percentage

This data helps editors understand which questions to add and which answers to rewrite.

Development Timeline

Total: 6 working days. Connecting full-text search via PostgreSQL tsvector — +1 day.