search-it-configuration

name	search-it-configuration
description	Search It backend configuration – search modes (LIKE/MATCH, AND/OR), text modes, highlight settings, additional index sources (DB columns, media, files), blacklist, plaintext conversion, similarity search (soundex/metaphone/Cologne phonetic), autocomplete/suggest setup, statistics. Use when the user configures search_it settings, adds index sources, or asks about search modes and options.

Search It – Configuration

All settings are managed in the REDAXO backend under Search It > Settings. They are stored in rex_config with addon key search_it.

Search mode settings

Logical mode

• AND – all search terms must appear (default, stricter)
• OR – any search term matches (broader results)

Text mode

• plain – search in plaintext index (recommended, ignores HTML)
• unmodified – search in original HTML content
• both – search in both (slower, more results)

Search mode (MySQL)

• LIKE – LIKE %term% queries, works always, slower on large indexes
• MATCH – MySQL FULLTEXT MATCH ... AGAINST, faster but requires FULLTEXT index and minimum word length (default 4 chars in MySQL)

Result settings

• Highlight type – how matched text is presented:
  • sentence – full sentence containing the match
  • paragraph – full paragraph
  • surroundtext – surrounding characters around the match
  • surroundtextsingle – surrounding characters, only first match
  • teaser – fixed-length teaser from start of content
  • array – returns array of matches (for custom rendering)
• Surround tags – HTML tags wrapping matches (default: <mark>, </mark>)
• Max highlighted chars – character limit for highlighted text
• Max results – hard limit on total results

Additional index sources

Under Settings > Additional sources, configure:

Database columns

Add any table.column pair. Common examples:

Table	Column	Purpose
rex_article	name	Article names in search results
rex_media	title	Media titles
rex_media	med_description	Media descriptions
Custom YForm table	name, description	Custom data records

After adding columns, trigger a full reindex.

File indexing

• Enable/disable file indexing
• Allowed file extensions (e.g. pdf,doc,docx,txt)
• Custom directories to scan (beyond media pool)

URL addon

If the URL addon is installed, enable URL indexing to include URL addon profile URLs in the search index.

Blacklist

Blacklisted words

Words that are stripped from search queries. Useful for very common terms that produce too many results.

Blacklisted articles

Specific article IDs excluded from indexing entirely.

Blacklisted categories

All articles in these categories (and optionally subcategories) are excluded from the index.

Similarity search

Enables "Did you mean?" suggestions when a search returns no results. Three phonetic algorithms available:

Mode	Algorithm	Best for
1	Soundex	English words
2	Metaphone	English words (better than Soundex)
4	Cologne Phonetic	German words
7	All combined	Mixed content

Enable in backend and rebuild the keyword index. The keyword index grows over time as users search – it learns from real queries.

Autocomplete / Suggest

1. Enable in Settings > Suggest
2. Configure minimum characters, maximum suggestions, search delay
3. The addon generates a JS snippet – embed it in your template before </body>
4. Search form must have class="search_it-form" and input name="search"

Generated JS code is available via:

rex_addon::get('search_it')->getProperty('suggest_js', '');

Alternatively, include the addon's CSS and JS assets manually:

<link rel="stylesheet" href="<?= rex_url::addonAssets('search_it', 'css/suggest.css') ?>">
<script src="<?= rex_url::addonAssets('search_it', 'js/suggest.js') ?>"></script>

Plaintext settings

Control how HTML content is converted to plaintext for indexing:

• CSS selectors to exclude – elements matching these selectors are removed before indexing (e.g. nav, footer, .search-exclude)
• Regex patterns – custom regex replacements applied during conversion
• Textile parsing – enable if content uses Textile markup

Statistics

When enabled, Search It tracks:

• Search terms used
• Frequency of each term
• Results count per term

View in backend under Search It > Statistics. Useful for understanding what visitors search for and improving content.

Indexing settings

• Host / Port – for socket-based article fetching during indexing (usually localhost:80)
• SSL verification – disable for self-signed certificates in development
• Logging – enable to log indexing errors to redaxo/data/addons/search_it/

Common pitfalls

• Setting search mode to MATCH but MySQL's ft_min_word_len is 4 – three-letter words are silently ignored. Check with SHOW VARIABLES LIKE 'ft_min_word_len'.
• Enabling Cologne Phonetic for an English-only site – produces poor suggestions. Use Metaphone instead.
• Configuring CSS selectors to exclude content but not reindexing – old content remains in the index.
• Enabling autocomplete but forgetting the search_it-form class on the form – JS doesn't bind.
• Blacklisting common words that are actually important search terms for the specific site.