komainu8

ktou

This plugin provides fast and rich full text search features based on <a href="http://mroonga.org/" rel="nofollow ugc">Mroonga</a>. Mroonga is a MySQL/MariaDB plugin. You don’t need to introduce a new server only for full text search. You can use existing MySQL/MariaDB server. It reduces maintainance cost.
Here are features:
<ul>
<li>Fast full text search</li>
<li>Google like query syntax such as <code>(KEYWORD1 OR KEYWORD2) -KEYWORD3</code></li>
<li>Sort by relevance</li>
</ul>
Here are features to be implemented:
<ul>
<li>Keyword highlight</li>
<li>Snippet</li>
<li>Related posts</li>
<li>Auto complete</li>
<li>Synonym</li>
<li>…</li>
</ul>

Mroonga

Fast and rich full text search features for WordPress!

ishitaka

Replaces site search with full-text search.
Replace the site search from LIKE search to Japanese full-text search (MySQL + Ngram parser or Mroonga engine + TokenMecab parser). This will significantly improve search performance.
Search for pure strings (plain text) without HTML tags. This will prevent HTML tags from being searched.
The data (index) for searching is stored in a dedicated table. It does not rewrite existing table structures or post data (posts table).
Searches for text in PDF, Word (doc, docx), Excel, and PowerPoint files. Secured PDF file are currently not supported.
It supports WordPress multisite.
<h4>Search string options</h4>
<ul>
<li><code>OR</code> (uppercase letter) – Combine searches. Example: foo OR bar</li>
<li><code>-</code> – Exclude words from the search. Example: foo -bar</li>
<li><code>""</code> – Search for an exact match. Example: “foo bar”</li>
<li><code>*</code> – Search by wildcard. Mroonga only. Example: foo*</li>
<li><code>()</code> – Grouping. Mroonga only. Example: (foo OR bar) baz</li>
</ul>
<h4>Operating environment</h4>
Requires MySQL 5.6 or later, or Mroonga engine.
Mroonga engine is strongly recommended. InnoDB engine performs significantly worse with large amounts of data.

Full-Text Search

Replaces site search with full-text search.

Mroonga vs Full-Text Search