Develop quick search prompts using PHP and Manticore Search

Use PHP and Manticore Search to develop a quick search prompt function

Introduction

Nowadays, with the development of the Internet, search engines have become one of the main tools for people to obtain information. Users usually use keywords to query content they are interested in, and search engines will return relevant results based on these keywords.

However, traditional search methods have some limitations. Users must enter complete keywords to get accurate results, which may increase the user's waiting time. To improve user experience, we can consider using the quick search prompts feature, which allows users to instantly get relevant search suggestions when entering keywords.

In this article, we will introduce how to use PHP and Manticore Search to implement quick search prompts.

Manticore Search Introduction

Manticore Search is an open source full-text search engine that is a fork of Sphinx Search and has higher performance and more features.

Manticore Search provides a function called "completion" for implementing search prompts. By building appropriate indexes and using appropriate query statements, we can achieve fast and accurate search suggestions.

Preparation

Before we begin, we need to install Manticore Search and ensure that PHP supports communicating with Manticore Search. We can install related extensions through the pecl command:

pecl install manticore

After successful installation, we need to enable this extension in the PHP extension configuration file. We can add the following lines in php.ini:

extension=manticore.so

Restart PHP for the changes to take effect.

Build index

Before we start, we need to prepare some data and create corresponding indexes. Suppose we have a table called "products" that contains product names and descriptions.

We can use the following code to create the index:

<?php

require_once 'ManticoreSearch.php';

$client = new ManticoreSearchClient([
    'host' => 'localhost',
    'port' => 9306,
]);

$client->connect();

$index = 'products';
$table = 'products';

$client->query("DELETE FROM $table WHERE 1");
$client->query("TRUNCATE RTINDEX $index");

$data = [
    ['id' => 1, 'name' => 'iPhone X', 'description' => 'A revolutionary new iPhone.'],
    ['id' => 2, 'name' => 'Samsung Galaxy S9', 'description' => 'The latest Samsung smartphone.'],
    ['id' => 3, 'name' => 'Google Pixel 3', 'description' => 'A powerful Android phone.'],
    // 添加更多产品数据...
];

$client->index($table, $data);
$client->query("ALTER RTINDEX $index ADD COLUMN completion TEXT");

$client->disconnect();
?>

In the above code, we first create a Client instance and then use connect Method to connect to the Manticore Search service. Next, we empty and rebuild the corresponding tables and indexes.

Next, we add some product data to the table. Make sure to add the actual product data to the $data array.

Finally, we use the ALTER RTINDEX command to add a column to the index for storing search hints.

Create search prompts

Now that we have the data and index ready, we can start creating search prompts. We can use the following code to achieve this:

<?php

require_once 'ManticoreSearch.php';

$client = new ManticoreSearchClient([
    'host' => 'localhost',
    'port' => 9306,
]);

$client->connect();

$index = 'products';
$table = 'products';

$keyword = isset($_GET['keyword']) ? $_GET['keyword'] : '';

$query = "SELECT * FROM $index WHERE MATCH(:keyword) OPTION field_weights=(name=10, description=5) LIMIT 10";
$params = [
    'keyword' => "@(name,description) $keyword*",
];

$result = $client->query($query, $params);

$suggestions = [];

if ($result !== false) {
    foreach ($result['hits'] as $hit) {
        $suggestions[] = $hit['attrs']['name'];
    }
}

$client->disconnect();

echo json_encode($suggestions);
?>

In the above code, we first create a Client instance and connect. Then, we define a $keyword variable to store the keywords entered by the user.

Next, we use the SELECT statement to execute the query. By including the keywords entered by the user as part of the query and using the MATCH clause to match content in the index, we can get results that match the keywords.

In this example, we have given a higher weight to the name field. You can adjust the field weight according to the actual situation.

Finally, we convert the search suggestions into JSON format and output it.

Conclusion

In this article, we learned how to use PHP and Manticore Search to implement quick search prompts. By building appropriate indexes and using the right query statements, we can enable users to instantly get relevant search suggestions as they enter keywords. This approach can greatly improve the user's search experience.

Due to space limitations, this article only provides a simple sample code, which you can modify and expand according to your own needs. I hope this article can help you further explore and use Manticore Search to build more powerful search capabilities.

The above is the detailed content of Develop quick search prompts using PHP and Manticore Search. For more information, please follow other related articles on the PHP Chinese website!