php+中文分词scws+sphinx+mysql打造千万级数据全文搜索

2018/01/05 15:39
阅读数 68

Sphinx是由俄罗斯人Andrew Aksyonoff开发的一个全文检索引擎。意图为其他应用提供高速、低空间占用、高结果 相关度的全文搜索功能。Sphinx可以非常容易的与SQL数据库和脚本语言集成。当前系统内置MySQL和PostgreSQL 数据库数据源的支持,也支持从标准输入读取特定格式 的XML数据。
Sphinx创建索引的速度为:创建100万条记录的索引只需3~4分钟,创建1000万条记录的索引可以在50分钟内完成,而只包含最新10万条记录的增量索引,重建一次只需几十秒。
Sphinx的特性如下:
a) 高速的建立索引(在当代CPU上,峰值性能可达到10 MB/秒);
b) 高性能的搜索(在2 – 4GB 的文本数据上,平均每次检索响应时间小于0.1秒);
c) 可处理海量数据(目前已知可以处理超过100 GB的文本数据, 在单一CPU的系统上可处理100 M 文档);
d) 提供了优秀的相关度算法,基于短语相似度和统计(BM25)的复合Ranking方法;
e) 支持分布式搜索;
f) 支持短语搜索
g) 提供文档摘要生成
h) 可作为MySQL的存储引擎提供搜索服务;
i) 支持布尔、短语、词语相似度等多种检索模式;
j) 文档支持多个全文检索字段(最大不超过32个);
k) 文档支持多个额外的属性信息(例如:分组信息,时间戳等);
l) 支持断词;
虽然mysql的MYISAM提供全文索引,但是性能却不敢让人恭维


开始搭建

系统环境:centos6.5+php5.6+apache+mysql

1、安装依赖包

1 yum -y install make gcc g++ gcc-c++ libtool autoconf automake imake php-devel mysql-devel libxml2-devel expat-devel

 

2、安装Sphinx

1 yum install expat expat-devel 
2 wget -c http://sphinxsearch.com/files/sphinx-2.0.7-release.tar.gz 
3 tar zxvf sphinx-2.0.7-release.tar.gz 
4 cd sphinx-2.0.7-release 
5 ./configure --prefix=/usr/local/sphinx --with-mysql --with-libexpat --enable-id64 
6 make && make install

 

3、安装libsphinxclient,php扩展用到

1 cd api/libsphinxclient 
2 ./configure --prefix=/usr/local/sphinx/libsphinxclient 
3 make && make install

 

4、安装Sphinx的PHP扩展:我的是5.6需装sphinx-1.3.3.tgz,如果是php5.4以下可sphinx-1.3.0.tgz

 
wget -c http://pecl.php.net/get/sphinx-1.3.3.tgz 
tar zxvf sphinx-1.3.3.tgz 
cd sphinx-1.3.3 
phpize 
./configure --with-sphinx=/usr/local/sphinx/libsphinxclient/ --with-php-config=/usr/bin/php-config 
make && make install 
成功后会提示: 
Installing shared extensions: /usr/lib64/php/modules/ 


echo "[Sphinx]" >> /etc/php.ini 
echo "extension = sphinx.so" >> /etc/php.ini 
#重启apache 
service httpd restart

 

5、创建测试数据
 

CREATE TABLE IF NOT EXISTS `items` ( 
`id` int(11) NOT NULL AUTO_INCREMENT, 
`title` varchar(255) NOT NULL, 
`content` text NOT NULL, 
`created` datetime NOT NULL, 
PRIMARY KEY (`id`) 
) ENGINE=InnoDB DEFAULT CHARSET=utf8 COMMENT='全文检索测试的数据表' AUTO_INCREMENT=11 ; 

INSERT INTO `items` (`id`, `title`, `content`, `created`) VALUES 
(1, 'linux mysql集群安装', 'MySQL Cluster 是MySQL 适合于分布式计算环境的高实用、可拓展、高性能、高冗余版本', '2016-09-07 00:00:00'), 
(2, 'mysql主从复制', 'mysql主从备份(复制)的基本原理 mysql支持单向、异步复制,复制过程中一个服务器充当主服务器,而一个或多个其它服务器充当从服务器', '2016-09-06 00:00:00'), 
(3, 'hello', 'can you search me?', '2016-09-05 00:00:00'), 
(4, 'mysql', 'mysql is the best database?', '2016-09-03 00:00:00'), 
(5, 'mysql索引', '关于MySQL索引的好处,如果正确合理设计并且使用索引的MySQL是一辆兰博基尼的话,那么没有设计和使用索引的MySQL就是一个人力三轮车', '2016-09-01 00:00:00'), 
(6, '集群', '关于MySQL索引的好处,如果正确合理设计并且使用索引的MySQL是一辆兰博基尼的话,那么没有设计和使用索引的MySQL就是一个人力三轮车', '0000-00-00 00:00:00'), 
(9, '复制原理', 'redis也有复制', '0000-00-00 00:00:00'), 
(10, 'redis集群', '集群技术是构建高性能网站架构的重要手段,试想在网站承受高并发访问压力的同时,还需要从海量数据中查询出满足条件的数据,并快速响应,我们必然想到的是将数据进行切片,把数据根据某种规则放入多个不同的服务器节点,来降低单节点服务器的压力', '0000-00-00 00:00:00'); 

CREATE TABLE IF NOT EXISTS `sph_counter` ( 
`counter_id` int(11) NOT NULL, 
`max_doc_id` int(11) NOT NULL, 
PRIMARY KEY (`counter_id`) 
) ENGINE=MyISAM DEFAULT CHARSET=utf8 COMMENT='增量索引标示的计数表';

 

以下采用"Main + Delta" ("主索引"+"增量索引")的索引策略,使用Sphinx自带的一元分词。
6、Sphinx配置:注意修改数据源配置信息
 

vi /usr/local/sphinx/etc/sphinx.conf 
source items { 
type = mysql 
sql_host = localhost 
sql_user = root 
sql_pass = 123456 
sql_db = sphinx_items 

sql_query_pre = SET NAMES utf8 
sql_query_pre = SET SESSION query_cache_type = OFF 
sql_query_pre = REPLACE INTO sph_counter SELECT 1, MAX(id) FROM items 

sql_query_range = SELECT MIN(id), MAX(id) FROM items \ 
WHERE id<=(SELECT max_doc_id FROM sph_counter WHERE counter_id=1) 
sql_range_step = 1000 
sql_ranged_throttle = 1000 

sql_query = SELECT id, title, content, created, 0 as deleted FROM items \ 
WHERE id<=(SELECT max_doc_id FROM sph_counter WHERE counter_id=1) \ 
AND id >= $start AND id <= $end 

sql_attr_timestamp = created 
sql_attr_bool = deleted 
} 

source items_delta : items { 
sql_query_pre = SET NAMES utf8 
sql_query_range = SELECT MIN(id), MAX(id) FROM items \ 
WHERE id > (SELECT max_doc_id FROM sph_counter WHERE counter_id=1) 
sql_query = SELECT id, title, content, created, 0 as deleted FROM items \ 
WHERE id>( SELECT max_doc_id FROM sph_counter WHERE counter_id=1 ) \ 
AND id >= $start AND id <= $end 
sql_query_post_index = set @max_doc_id :=(SELECT max_doc_id FROM sph_counter WHERE counter_id=1) 
sql_query_post_index = REPLACE INTO sph_counter SELECT 2, IF($maxid, $maxid, @max_doc_id) 
} 
#主索引 
index items { 
source = items 
path = /usr/local/sphinx/var/data/items 
docinfo = extern 
morphology = none 
min_word_len = 1 
min_prefix_len = 0 
html_strip = 1 
html_remove_elements = style, script 
ngram_len = 1 
ngram_chars = U+3000..U+2FA1F 
charset_type = utf-8 
charset_table = 0..9, A..Z->a..z, _, a..z, U+410..U+42F->U+430..U+44F, U+430..U+44F 
preopen = 1 
min_infix_len = 1 
} 

#增量索引 
index items_delta : items { 
source = items_delta 
path = /usr/local/sphinx/var/data/items-delta 
} 

#分布式索引 
index master { 
type = distributed 
local = items 
local = items_delta 
} 

indexer { 
mem_limit = 256M 
} 

searchd { 
listen = 9312 
listen = 9306:mysql41 #Used for SphinxQL 
log = /usr/local/sphinx/var/log/searchd.log 
query_log = /usr/local/sphinx/var/log/query.log 
compat_sphinxql_magics = 0 
attr_flush_period = 600 
mva_updates_pool = 16M 
read_timeout = 5 
max_children = 0 
dist_threads = 2 
pid_file = /usr/local/sphinx/var/log/searchd.pid 
max_matches = 1000 
seamless_rotate = 1 
preopen_indexes = 1 
unlink_old = 1 
workers = threads # for RT to work 
binlog_path = /usr/local/sphinx/var/data 

} 

保存退出

7、Sphinx创建索引

#第一次需重建索引: 
[root@localhost bin]# ./indexer -c /usr/local/sphinx/etc/sphinx.conf --all 
Sphinx 2.0.7-id64-release (r3759) 
Copyright (c) 2001-2012, Andrew Aksyonoff 
Copyright (c) 2008-2012, Sphinx Technologies Inc (http://sphinxsearch.com) 

using config file '/usr/local/sphinx/etc/sphinx.conf'... 
indexing index 'items'... 
collected 8 docs, 0.0 MB 
sorted 0.0 Mhits, 100.0% done 
total 8 docs, 1121 bytes 
total 1.017 sec, 1101 bytes/sec, 7.86 docs/sec 
indexing index 'items_delta'... 
collected 0 docs, 0.0 MB 
total 0 docs, 0 bytes 
total 1.007 sec, 0 bytes/sec, 0.00 docs/sec 
skipping non-plain index 'master'... 
total 4 reads, 0.000 sec, 0.7 kb/call avg, 0.0 msec/call avg 
total 14 writes, 0.001 sec, 0.5 kb/call avg, 0.1 msec/call avg 
#启动sphinx 
[root@localhost bin]# ./searchd -c /usr/local/sphinx/etc/sphinx.conf 
Sphinx 2.0.7-id64-release (r3759) 
Copyright (c) 2001-2012, Andrew Aksyonoff 
Copyright (c) 2008-2012, Sphinx Technologies Inc (http://sphinxsearch.com) 

using config file '/usr/local/sphinx/etc/sphinx.conf'... 
listening on all interfaces, port=9312 
listening on all interfaces, port=9306 
precaching index 'items' 
precaching index 'items_delta' 
rotating index 'items_delta': success 
precached 2 indexes in 0.012 sec

 


#查看进程

[root@localhost bin]# ps -ef | grep searchd 
root 30431 1 0 23:59 ? 00:00:00 ./searchd -c /usr/local/sphinx/etc/sphinx.conf 
root 30432 30431 0 23:59 ? 00:00:00 ./searchd -c /usr/local/sphinx/etc/sphinx.conf 
root 30437 1490 0 23:59 pts/0 00:00:00 grep searchd 
#停止Searchd: 
./searchd -c /usr/local/sphinx/etc/sphinx.conf --stop 
#查看Searchd状态: 
./searchd -c /usr/local/sphinx/etc/sphinx.conf --status

 

索引更新及使用说明
"增量索引"每N分钟更新一次.通常在每天晚上低负载的时进行一次索引合并,同时重新建立"增量索引"。当然"主索引"数据不多的话,也可以直接重新建立"主索引"。
API搜索的时,同时使用"主索引"和"增量索引",这样可以获得准实时的搜索数据.本文的Sphinx配置将"主索引"和"增量索引"放到分布式索引master中,因此只需查询分布式索引"master"即可获得全部匹配数据(包括最新数据)。

索引的更新与合并的操作可以放到cron job完成:
 

crontab -e 
*/1 * * * * /usr/local/sphinx/shell/delta_index_update.sh 
0 3 * * * /usr/local/sphinx/shell/merge_daily_index.sh 
crontab -l

cron job所用的shell脚本例子:

delta_index_update.sh:
[php] view plain copy
#!/bin/bash 
/usr/local/sphinx/bin/indexer -c /usr/local/sphinx/etc/sphinx.conf --rotate items_delta > /dev/null 2>&1 
merge_daily_index.sh:
[php] view plain copy
#!/bin/bash 
indexer=`which indexer` 
mysql=`which mysql` 

QUERY="use sphinx_items;select max_doc_id from sph_counter where counter_id = 2 limit 1;" 
index_counter=$($mysql -h192.168.1.198 -uroot -p123456 -sN -e "$QUERY") 

#merge "main + delta" indexes 
$indexer -c /usr/local/sphinx/etc/sphinx.conf --rotate --merge items items_delta --merge-dst-range deleted 0 0 >> /usr/local/sphinx/var/index_merge.log 2>&1 

if [ "$?" -eq 0 ]; then 
##update sphinx counter 
if [ ! -z $index_counter ]; then 
$mysql -h192.168.1.198 -uroot -p123456 -Dsphinx_items -e "REPLACE INTO sph_counter VALUES (1, '$index_counter')" 
fi 
##rebuild delta index to avoid confusion with main index 
$indexer -c /usr/local/sphinx/etc/sphinx.conf --rotate items_delta >> /usr/local/sphinx/var/rebuild_deltaindex.log 2>&1 
fi

 

8、php中文分词scws安装:注意扩展的版本和php的版本

wget -c http://www.xunsearch.com/scws/down/scws-1.2.3.tar.bz2 
tar jxvf scws-1.2.3.tar.bz2 
cd scws-1.2.3 
./configure --prefix=/usr/local/scws 
make && make install

 

9、scws的PHP扩展安装:

cd ./phpext 
phpize 
./configure 
make && make install 
echo "[scws]" >> /etc/php.ini 
echo "extension = scws.so" >> /etc/php.ini 
echo "scws.default.charset = utf-8" >> /etc/php.ini 
echo "scws.default.fpath = /usr/local/scws/etc/" >> /etc/php.ini

10、词库安装:

wget http://www.xunsearch.com/scws/down/scws-dict-chs-utf8.tar.bz2 
tar jxvf scws-dict-chs-utf8.tar.bz2 -C /usr/local/scws/etc/ 
chown www:www /usr/local/scws/etc/dict.utf8.xdb

11、php使用Sphinx+scws测试例子
在Sphinx源码API中,有好几种语言的API调用.其中有一个是sphinxapi.php。
不过以下的测试使用的是Sphinx的PHP扩展.具体安装见本文开头的Sphinx安装部分。
测试用的搜索类Search.php:注意修改getDBConnection()信息为自己的

<?php 
class Search { 
/** 
* @var SphinxClient 
**/ 
protected $client; 
/** 
* @var string 
**/ 
protected $keywords; 
/** 
* @var resource 
**/ 
private static $dbconnection = null; 

/** 
* Constructor 
**/ 
public function __construct($options = array()) { 
$defaults = array( 
'query_mode' => SPH_MATCH_EXTENDED2, 
'sort_mode' => SPH_SORT_EXTENDED, 
'ranking_mode' => SPH_RANK_PROXIMITY_BM25, 
'field_weights' => array(), 
'max_matches' => 1000, 
'snippet_enabled' => true, 
'snippet_index' => 'items', 
'snippet_fields' => array(), 
); 
$this->options = array_merge($defaults, $options); 
$this->client = new SphinxClient(); 
//$this->client->setServer("192.168.1.198", 9312); 
$this->client->setMatchMode($this->options['query_mode']); 
if ($this->options['field_weights'] !== array()) { 
$this->client->setFieldWeights($this->options['field_weights']); 
} 
/* 
if ( in_array($this->options['query_mode'], [SPH_MATCH_EXTENDED2,SPH_MATCH_EXTENDED]) ) { 
$this->client->setRankingMode($this->options['ranking_mode']); 
} 
*/ 
} 

/** 
* Query 
* 
* @param string $keywords 
* @param integer $offset 
* @param integer $limit 
* @param string $index 
* @return array 
**/ 
public function query($keywords, $offset = 0, $limit = 10, $index = '*') { 
$this->keywords = $keywords; 
$max_matches = $limit > $this->options['max_matches'] ? $limit : $this->options['max_matches']; 
$this->client->setLimits($offset, $limit, $max_matches); 
$query_results = $this->client->query($keywords, $index); 

if ($query_results === false) { 
$this->log('error:' . $this->client->getLastError()); 
} 

$res = []; 
if ( empty($query_results['matches']) ) { 
return $res; 
} 
$res['total'] = $query_results['total']; 
$res['total_found'] = $query_results['total_found']; 
$res['time'] = $query_results['time']; 
$doc_ids = array_keys($query_results['matches']); 
unset($query_results); 
$res['data'] = $this->fetch_data($doc_ids); 
if ($this->options['snippet_enabled']) { 
$this->buildExcerptRows($res['data']); 
} 

return $res; 
} 

/** 
* custom sorting 
* 
* @param string $sortBy 
* @param int $mode 
* @return bool 
**/ 
public function setSortBy($sortBy = '', $mode = 0) { 
if ($sortBy) { 
$mode = $mode ?: $this->options['sort_mode']; 
$this->client->setSortMode($mode, $sortBy); 
} else { 
$this->client->setSortMode(SPH_SORT_RELEVANCE); 
} 
} 

/** 
* fetch data based on matched doc_ids 
* 
* @param array $doc_ids 
* @return array 
**/ 
protected function fetch_data($doc_ids) { 
$ids = implode(',', $doc_ids); 
$queries = self::getDBConnection()->query("SELECT * FROM items WHERE id in ($ids)", PDO::FETCH_ASSOC); 
return iterator_to_array($queries); 
} 

/** 
* build excerpts for data 
* 
* @param array $rows 
* @return array 
**/ 
protected function buildExcerptRows(&$rows) { 
$options = array( 
'before_match' => '<b style="color:red">', 
'after_match' => '</b>', 
'chunk_separator' => '...', 
'limit' => 256, 
'around' => 3, 
'exact_phrase' => false, 
'single_passage' => true, 
'limit_words' => 5, 
); 
$scount = count($this->options['snippet_fields']); 
foreach ($rows as &$row) { 
foreach ($row as $fk => $item) { 
if (!is_string($item) || ($scount && !in_array($fk, $this->options['snippet_fields'])) ) continue; 
$item = preg_replace('/[\r\t\n]+/', '', strip_tags($item)); 
$res = $this->client->buildExcerpts(array($item), $this->options['snippet_index'], $this->keywords, $options); 
$row[$fk] = $res === false ? $item : $res[0]; 
} 
} 
return $rows; 
} 

/** 
* database connection 
* 
* @return resource 
**/ 
private static function getDBConnection() { 
$dsn = 'mysql:host=192.168.1.198;dbname=sphinx_items'; 
$user = 'root'; 
$pass = '123456'; 
if (!self::$dbconnection) { 
try { 
self::$dbconnection = new PDO($dsn, $user, $pass); 
} catch (PDOException $e) { 
die('Connection failed: ' . $e->getMessage()); 
} 
} 
return self::$dbconnection; 
} 

/** 
* Chinese words segmentation 
* 
**/ 
public function wordSplit($keywords) { 
$fpath = ini_get('scws.default.fpath'); 
$so = scws_new(); 
$so->set_charset('utf-8'); 
$so->add_dict($fpath . '/dict.utf8.xdb'); 
//$so->add_dict($fpath .'/custom_dict.txt', SCWS_XDICT_TXT); 
$so->set_rule($fpath . '/rules.utf8.ini'); 
$so->set_ignore(true); 
$so->set_multi(false); 
$so->set_duality(false); 
$so->send_text($keywords); 
$words = []; 
$results = $so->get_result(); 
foreach ($results as $res) { 
$words[] = '(' . $res['word'] . ')'; 
} 
$words[] = '(' . $keywords . ')'; 
return join('|', $words); 
} 

/** 
* get current sphinx client 
* 
* @return resource 
**/ 
public function getClient() { 
return $this->client; 
} 
/** 
* log error 
**/ 
public function log($msg) { 
// log errors here 
//echo $msg; 
} 
/** 
* magic methods 
**/ 
public function __call($method, $args) { 
$rc = new ReflectionClass('SphinxClient'); 
if ( !$rc->hasMethod($method) ) { 
throw new Exception('invalid method :' . $method); 
} 
return call_user_func_array(array($this->client, $method), $args); 
} 
}

 

测试文件test.php:

<?php 
require('Search.php'); 
$s = new Search([ 
'snippet_fields' => ['title', 'content'], 
'field_weights' => ['title' => 20, 'content' => 10], 
]); 
$s->setSortMode(SPH_SORT_EXTENDED, 'created desc,@weight desc'); 
//$s->setSortBy('created desc,@weight desc'); 
$words = $s->wordSplit("mysql集群");//先分词 结果:(mysql)|(mysql集群) 
//print_r($words);exit; 
$res = $s->query($words, 0, 10, 'master'); 
echo '<pre/>';print_r($res);

测试结果:

12、SphinxQL测试
要使用SphinxQL需要在Searchd的配置里面增加相应的监听端口(参考上文配置)。

[root@localhost bin]# mysql -h127.0.0.1 -P9306 -uroot -p 
Enter password: 
Welcome to the MySQL monitor. Commands end with ; or \g. 
Your MySQL connection id is 1 
Server version: 2.0.7-id64-release (r3759) 

Copyright (c) 2000, 2016, Oracle and/or its affiliates. All rights reserved. 

Oracle is a registered trademark of Oracle Corporation and/or its 
affiliates. Other names may be trademarks of their respective 
owners. 

Type 'help;' or '\h' for help. Type '\c' to clear the current input statement. 

mysql> show global variables; 
+----------------------+---------+ 
| Variable_name | Value | 
+----------------------+---------+ 
| autocommit | 1 | 
| collation_connection | libc_ci | 
| query_log_format | plain | 
| log_level | info | 
+----------------------+---------+ 
4 rows in set (0.00 sec) 

mysql> desc items; 
+---------+-----------+ 
| Field | Type | 
+---------+-----------+ 
| id | bigint | 
| title | field | 
| content | field | 
| created | timestamp | 
| deleted | bool | 
+---------+-----------+ 
5 rows in set (0.00 sec) 

mysql> select * from master where match ('mysql集群') limit 10; 
+------+---------+---------+ 
| id | created | deleted | 
+------+---------+---------+ 
| 1 | 2016 | 0 | 
| 6 | 0 | 0 | 
+------+---------+---------+ 
2 rows in set (0.00 sec) 

mysql> show meta; 
+---------------+-------+ 
| Variable_name | Value | 
+---------------+-------+ 
| total | 2 | 
| total_found | 2 | 
| time | 0.006 | 
| keyword[0] | mysql | 
| docs[0] | 5 | 
| hits[0] | 15 | 
| keyword[1] | 集 | 
| docs[1] | 3 | 
| hits[1] | 4 | 
| keyword[2] | 群 | 
| docs[2] | 3 | 
| hits[2] | 4 | 
+---------------+-------+ 
12 rows in set (0.00 sec) 

mysql>

 

展开阅读全文
打赏
0
0 收藏
分享
加载中
更多评论
打赏
0 评论
0 收藏
0
分享
返回顶部
顶部