Detailed example explanation of PHP+Sphinx+Mysql development of search engine

When everyone hears about search engines, they will find it difficult to write and have no idea at all. In fact, PHP can also be used for search engine development, but PHP needs to be combined with sphinx and mysql to develop the search engine we want. I want to know about PHP How to carry out search engine development! Let’s take a look! !

First we download the Sphinx tool, download address official website download address: www.sphinxsearch.com, find sphinx-2.2.10-release-win64.zip, download this for 64-bit, unzip it to us Under the PHP running directory, it is convenient to view the results on subsequent web pages.

sphinx introduction:

Sphinx is the abbreviation of SQL Phrase Index (query phrase index). Sphinx is a full-text search engine based on SQL. The API interfaces it provides include :PHP, Python, Perl, Ruby, java, etc. At the same time, an engine plug-in SphinxSE is designed for MySQL, which is a distributed full-text retrieval system.
Advantages:
High-speed indexing can reach 10M/s
High-performance search (text data in 2-4G On average, the average response time for each retrieval is less than 0.1 seconds)
Can handle massive amounts of data (currently known to be able to process 100G of text data, and 100M of documents on a single CPU system)
Provides excellent correlation algorithm, composite ranking method based on phrase similarity and statistical BM2
Supports distributed search
Provide document fragment generation function
Can be used as a Mysql storage engine to provide search services
Support Boolean, phrase, word similarity and other search modes
Disadvantages:
Must have a primary key
The primary key must be an integer
Not responsible for data storage
The configuration is not flexible

The sphinx structure after decompression is as shown in the figure:

The following is our process For related configuration, see sphinx-min.conf.in in the picture, copy it to our bin directory for easy use and change the name to sphinx.conf,

Modify the content inside:

source src1
{
	type			= mysql

	sql_host		= localhost #主机地址
	sql_user		= root#帐号
	sql_pass		=     #密码
	sql_db			= sphinx  #数据库
	sql_port		= 3306	# 数据库端口 3306
	sql_query		= SELECT id, name, age FROM users #查询语句
	sql_attr_uint		= group_id
	sql_attr_timestamp	= date_added
	sql_query_pre = set names utf8   #数据库编码
}


index test1
{
	source			= src1
	path			= D:/myapaphe/www/sphinx/data #这个一定要配置
	charset_type = utf-8 #指定编码
	ngram_len = 1        #要找中文需指定为1.
	ngram_chars = U+3000..U+2FA1F
	
}

indexer
{
	mem_limit		= 128M
}
searchd
{
	listen			= 9312
	listen			= 9306:mysql41
	log			= D:\myapaphe\www\sphinx\log\searchd.log  #进程日志
	query_log		= D:\myapaphe\www\sphinx\log\query.log    #查询日志

	read_timeout		= 5
	max_children		= 30
	pid_file		= D:\myapaphe\www\sphinx\log\searchd.pid 
	seamless_rotate		= 1
	preopen_indexes		= 1
	unlink_old		= 1
	workers			= threads # for RT to work
	binlog_path		= D:\myapaphe\www\sphinx\data
}

The above must be configured, and the path must match your own path.

Next generate the query index:

Install searchd service:

Next load the configuration file:

SetServer('localhost',9312);
$result = $s->Query('高七');
echo '
';
print_r($result);


Garbled characters are because cmd defaults to gbk encoding. Let’s put it in the browser to view: 



We see that sphinx does not find the complete result but returns the ID to us, allowing us to check the data based on the ID. 
The following is a query time comparison: 


The time I tested on more than 40,000 pieces of data was 0.001s. Let’s take a look at mysql How long does the query take: 


 We see that it takes 0.04s, there is not much data, and the result is not that obvious, but the gap of 0.039s is not small. 
This completes the integration of sphinx, I hope it can help everyone. 

related suggestion:

php Detailed explanation of calling existing search engines
php Function code to determine whether the visitor is a search engine spider
The above is the detailed content of Detailed example explanation of PHP+Sphinx+Mysql development of search engine. For more information, please follow other related articles on the PHP Chinese website!