How to build a full-text search engine using PHP and Elasticsearch

How to build a full-text search engine using PHP and Elasticsearch

Full-text search engines play an important role in the modern Internet, allowing users to quickly find information that meets their needs. A good full-text search engine not only needs to search quickly, but also needs to provide high-quality search results. This article will teach you how to build a full-text search engine using PHP and Elasticsearch.

What is Elasticsearch?

Elasticsearch is an open source search engine based on the Lucene search engine library. It provides a distributed, multi-tenant full-text search engine and is capable of automatically processing large-scale data. Elasticsearch can search and analyze data quickly and supports real-time search, which means that as the data is updated, Elasticsearch can return search results from new data within milliseconds.

Elasticsearch supports a variety of data types, including text, numerical values, dates, geographical locations, etc. By using Elasticsearch, we can quickly build a full-text search engine and customize it according to different needs.

Connecting to Elasticsearch using PHP

To connect to Elasticsearch using PHP, we need to install the Elasticsearch PHP client library. The easiest way to install this library is to use Composer, which can be installed with the following command:

composer require elasticsearch/elasticsearch

After the installation is complete, reference this library in the code:

require 'vendor/autoload.php';

use ElasticsearchClientBuilder;

Here we use the Elasticsearch PHP client Library's ClientBuilder class to connect to Elasticsearch.

$config = [

'hosts' => ['localhost:9200']

];

$client = ClientBuilder::create()->setHosts($config['hosts'])-> build();

Here we use the ClientBuilder class to create an Elasticsearch client and connect using the host name and port of the cluster.

Create Index

In a full-text search engine, data needs to be stored in the index, not in the database. To create an index, we first need to define the structure of the data and the settings for the index. This structure is called mapping.

For this example, let's assume we want to create a search engine to search for articles. Articles have fields such as title, author, publication date, and text. We can define mapping using the following code:

$params = [

'index' => 'articles',
'body' => [
    'mappings' => [
        'article' => [
            'properties' => [
                'title' => ['type' => 'text'],
                'author' => ['type' => 'text'],
                'publish_date' => ['type' => 'date'],
                'body' => ['type' => 'text']
            ]
        ]
    ]
]

];

$response = $client->indices()->create($ params);

Here we define an index named "articles" and define the mapping of articles, including the title, author, publication date and text of the article. This will create an index called "articles" containing a type called "article" which contains the fields we defined.

After creating the index, we can start adding data to the index.

Add data

To add data, we write the data to Elasticsearch by calling the index method:

$params = [

'index' => 'articles',
'type' => 'article',
'id' => 1,
'body' => [
    'title' => '如何使用PHP和Elasticsearch构建全文搜索引擎',
    'author' => 'John Doe',
    'publish_date' => '2020-01-01',
    'body' => '全文搜索引擎在现代互联网中扮演着重要的角色……'
]

];

$client->index($params);

Here we specify the "articles" index and "article" type to be indexed, and use unique ID. When indexing data, we write the data into Elasticsearch, index it, and make it searchable.

Query data

In order to query data, we need to construct a query request and send it to the Elasticsearch server. We can use match query, which will match keywords in fields. For example:

$query = [

'match' => [
    'title' => 'Elasticsearch'
]

];

$params = [

'index' => 'articles',
'type' => 'article',
'body' => [
    'query' => $query
]

];

$response = $ client->search($params);

Here we query all documents of type "article" in the "articles" index that contain the "title" field, and the "title" field contains the keyword "Elasticsearch".

Elasticsearch also supports many other types of queries, such as Boolean queries, range queries, prefix queries, etc. Understanding all query types and how to use them can make search engines more adaptable to different types of data.

Conclusion

In this article, we learned how to build a full-text search engine using PHP and Elasticsearch. We first installed the Elasticsearch PHP client library and used it to connect to the Elasticsearch server. Then, we create an index called "articles" and define the mapping of the articles. After that, we added some data and ran a query. Finally, we learned that Elasticsearch supports many query types and learned how to use them so that we can build a more powerful full-text search engine suitable for different data types.

The above is the detailed content of How to build a full-text search engine using PHP and Elasticsearch. For more information, please follow other related articles on the PHP Chinese website!