您现在的位置是:首页 > cms教程 > Discuz教程Discuz教程

discuz的sphinx全文检索搜索引擎方案

关戚栾2025-07-16Discuz教程已有人查阅

导读sphinx最主要的就是这个配置文件,当然在增量索引部分可以写一个脚本放到crontab里面来定时跑下面介绍下在discuz中sphinx的php调用部分,sphinx的接口采用PHP的扩展

#
# linuxTone full index search configure file
#
source lt_posts
{
type = mysql
sql_host = 127.0.0.1
sql_user = root
sql_pass =
sql_db = lt_bbs
sql_port = 3306
sql_query_pre = SET NAMES utf8
sql_query = SELECT pid,tid,fid,dateline,subject,message,author FROM cdb_posts where first=1 #此处是基于posts表来做索引的,这样的目的是可以同时检索到subject,message,author 三个字段的值
sql_attr_uint = fid
sql_attr_timestamp = dateline
sql_query_info = SELECT * FROM cdb_posts WHERE pid=$id
}
index lt_posts
{
source = lt_posts
path = /data/sphinx/data/lt_posts
docinfo = extern
mlock = 0
morphology = none
min_word_len = 2
html_strip = 1
charset_dictpath = /usr/local/mmseg-3.2.13/etc/
charset_type = zh_cn.utf-8
ngram_len = 0
}
########## 增量索引 ##################
source delta
{
type = mysql
sql_host = 127.0.0.1
sql_user = root
sql_pass =
sql_db = lt_bbs
sql_port = 3306 # optional, default is 3306
sql_query_pre = SET NAMES utf8
sql_query = SELECT pid,tid,fid,dateline,subject,message,author FROM cdb_posts where first=1 and dateline > unix_timestamp()-3600*10 #增量索引采用当前时间戳减去一个需要间隔的时间来新建新增的数据索引
sql_attr_uint = fid
sql_attr_timestamp = dateline
sql_query_info = SELECT * FROM cdb_posts WHERE pid=$id
}
index delta
{
source = delta
path = /data/sphinx/data/lt_delta
docinfo = extern
mlock = 0
morphology = none
min_word_len = 2
html_strip = 1
charset_dictpath = /usr/local/mmseg-3.2.13/etc/
charset_type = zh_cn.utf-8
ngram_len = 0
}
indexer
{
mem_limit = 32M
}
searchd
{
port = 9312
log = /data/sphinx/var/log/searchd.log
query_log = /data/sphinx/var/log/query.log
read_timeout = 5
max_children = 30
pid_file = /data/sphinx/var/log/searchd.pid
max_matches = 10000
seamless_rotate = 1
preopen_indexes = 0
unlink_old = 1
}
sphinx最主要的就是这个配置文件,当然在增量索引部分可以写一个脚本放到crontab里面来定时跑下面介绍下在discuz中sphinx的php调用部分,sphinx的接口采用PHP的扩展,可以通过pecl或者来安装
<?php
/**
*全文搜索服务
*/
define('IN_DISCUZ', true);
require_once './include/common.inc.php';
$q = isset($_GET['q']) && !empty($_GET['q']) ? $_GET['q'] : '';
$q = str_replace(array('<', '>', ' ', '\'', ','), array('', '', ' ', '', ''), strip_tags($q));
$page = isset($_GET['page']) && intval($_GET['page']) > 0 ? intval($_GET['page']) : 1;
$perNum = 20;
$offset = ($page - 1) * $perNum;
$search = new SphinxClient();
$search -> setServer('127.0.0.1', 9312);
$search -> setConnectTimeout(2);
$search -> setArrayResult(true);
$search -> setMatchMode(SPH_MATCH_ANY);
$search -> setRankingMode(SPH_RANK_PROXIMITY_BM25);
$search -> setSortMode(SPH_SORT_EXTENDED, '@relevance desc,@weight desc');
$search -> setLimits($offset, $perNum);
$search -> setFieldWeights(array('subject' => 2000, 'message' => 0));
$rs = array();
$query_totals = $query_time = 0;
if (!empty($q)) {
$rs = $search -> Query($q, "*");
$pages = ceil($rs['total'] / $perNum);
$query_totals = $rs['total_found'];
$query_time = $rs['time'];
}
$data = $title = $content = array();
if (!empty($rs) && $page <= $pages) {
$pids = array();
foreach($rs['matches'] as $v) {
$pids[] = $v['id'];
}
$pid = implode(',', $pids);
$sql = "select pid,tid,author,authorid,subject,message,dateline from cdb_posts where pid IN($pid) and status ='0' and invisible='0'";
$query = $db -> query($sql);
while ($row = $db -> fetch_array($query)) {
$data[] = $row;
$title[] = $row['subject'];
$content[] = preg_replace('/\[[\/]?(b|img|url|color|s|hr|p|list|i|align|email|u|font|code|hide|table|tr|td|th|attach|list|indent|float).*\]/', '', strip_tags($row['message']));
}
// 搜索词高亮
$opts = array();
$opts['before_match'] = '<em>';
$opts['after_match'] = '</em>';
$title = $search -> BuildExcerpts($title, 'lt_posts', $q, $opts);
$content = $search -> BuildExcerpts($content, 'lt_posts', $q, $opts);
foreach($data as $k => $v) {
$data[$k]['subject'] = $title[$k];
$data[$k]['message'] = $content[$k];
}
$url = "s.php?q=" . urlencode($q);
$multipage = multi($rs['total'], $perNum, $page, $url);
}
include template("lt_search");
?>
跑主索引的shell脚本search-index.sh
#!/bin/bash
#
# The BBS search exec full index
#
/usr/local/csft-3.2.13/bin/indexer -c /usr/local/csft-3.2.13/etc/lt_posts.conf --rotate lt_posts >> /data/sphinx/var/`date "+%Y-%m-%d-%H"`.log
跑增量索引
/usr/local/csft-3.2.13/bin/indexer -c /usr/local/csft-3.2.13/etc/lt_posts.conf --rotate delta
合并主索引和增量索引
#/usr/local/csft-3.2.13/bin/indexer --config /usr/local/csft-3.2.13/etc/lt_posts.conf --rotate --merge lt_posts delta
1 <?php
2
3 /*
4 [UCenter] (C)2001-2099 Comsenz Inc.
5 This is NOT a freeware, use is subject to license terms
6
7 $Id: db.class.php 1059 2011-03-01 07:25:09Z monkey $
8 */
9
10
11 class ucclient_db {
12 var $querynum = 0;
13 var $link;
14 var $histories;
15
16 var $dbhost;
17 var $dbuser;
18 var $dbpw;
19 var $dbcharset;
20 var $pconnect;
21 var $tablepre;
22 var $time;
23
24 var $goneaway = 5;
25
26 function connect($dbhost, $dbuser, $dbpw, $dbname = '', $dbcharset = '', $pconnect = 0, $tablepre='', $time = 0) {
27 $this->dbhost = $dbhost;
28 $this->dbuser = $dbuser;
29 $this->dbpw = $dbpw;
30 $this->dbname = $dbname;
31 $this->dbcharset = $dbcharset;
32 $this->pconnect = $pconnect;
33 $this->tablepre = $tablepre;
34 $this->time = $time;
35
36 if($pconnect) {
37 if(!$this->link = mysql_pconnect($dbhost, $dbuser, $dbpw)) {
38 $this->halt('Can not connect to MySQL server');
39 }
40 } else {
41 if(!$this->link = mysql_connect($dbhost, $dbuser, $dbpw)) {
42 $this->halt('Can not connect to MySQL server');
43 }
44 }
45
46 if($this->version() > '4.1') {
47 if($dbcharset) {
48 mysql_query("SET character_set_connection=".$dbcharset.", character_set_results=".$dbcharset.", character_set_client=binary", $this->link);
49 }
50
51 if($this->version() > '5.0.1') {
52 mysql_query("SET sql_mode=''", $this->link);
53 }
54 }
55
56 if($dbname) {
57 mysql_select_db($dbname, $this->link);
58 }
59
60 }
61
62 function fetch_array($query, $result_type = MYSQL_ASSOC) {
63 return mysql_fetch_array($query, $result_type);
64 }
65
66 function result_first($sql) {
67 $query = $this->query($sql);
68 return $this->result($query, 0);
69 }
70
71 function fetch_first($sql) {
72 $query = $this->query($sql);
73 return $this->fetch_array($query);
74 }
75
76 function fetch_all($sql, $id = '') {
77 $arr = array();
78 $query = $this->query($sql);
79 while($data = $this->fetch_array($query)) {
80 $id ? $arr[$data[$id]] = $data : $arr[] = $data;
81 }
82 return $arr;
83 }
84
85 function cache_gc() {
86 $this->query("DELETE FROM {$this->tablepre}sqlcaches WHERE expiry<$this->time");
87 }
88
89 function query($sql, $type = '', $cachetime = FALSE) {
90 $func = $type == 'UNBUFFERED' && @function_exists('mysql_unbuffered_query') ? 'mysql_unbuffered_query' : 'mysql_query';
91 if(!($query = $func($sql, $this->link)) && $type != 'SILENT') {
92 $this->halt('MySQL Query Error', $sql);
93 }
94 $this->querynum++;
95 $this->histories[] = $sql;
96 return $query;
97 }
98
99 function affected_rows() {
100 return mysql_affected_rows($this->link);
101 }
102
103 function error() {
104 return (($this->link) ? mysql_error($this->link) : mysql_error());
105 }
106
107 function errno() {
108 return intval(($this->link) ? mysql_errno($this->link) : mysql_errno());
109 }
110
111 function result($query, $row) {
112 $query = @mysql_result($query, $row);
113 return $query;
114 }
115
116 function num_rows($query) {
117 $query = mysql_num_rows($query);
118 return $query;
119 }
120
121 function num_fields($query) {
122 return mysql_num_fields($query);
123 }
124
125 function free_result($query) {
126 return mysql_free_result($query);
127 }
128
129 function insert_id() {
130 return ($id = mysql_insert_id($this->link)) >= 0 ? $id : $this->result($this->query("SELECT last_insert_id()"), 0);
131 }
132
133 function fetch_row($query) {
134 $query = mysql_fetch_row($query);
135 return $query;
136 }
137
138 function fetch_fields($query) {
139 return mysql_fetch_field($query);
140 }
141
142 function version() {
143 return mysql_get_server_info($this->link);
144 }
145
146 function close() {
147 return mysql_close($this->link);
148 }
149
150 function halt($message = '', $sql = '') {
151 $error = mysql_error();
152 $errorno = mysql_errno();
153 if($errorno == 2006 && $this->goneaway-- > 0) {
154 $this->connect($this->dbhost, $this->dbuser, $this->dbpw, $this->dbname, $this->dbcharset, $this->pconnect, $this->tablepre, $this->time);
155 $this->query($sql);
156 } else {
157 $s = '';
158 if($message) {
159 $s = "<b>UCenter info:</b> $message<br />";
160 }
161 if($sql) {
162 $s .= '<b>SQL:</b>'.htmlspecialchars($sql).'<br />';
163 }
164 $s .= '<b>Error:</b>'.$error.'<br />';
165 $s .= '<b>Errno:</b>'.$errorno.'<br />';
166 $s = str_replace(UC_DBTABLEPRE, '[Table]', $s);
167 exit($s);
168 }
169 }
170 }
171
172 ?>

本文标签:

很赞哦! ()

相关源码

  • (自适应)挖掘机大型采矿设备pbootcms网站源码下载本模板基于PbootCMS系统开发,专为重型机械设备行业设计,特别适合挖掘机、采矿设备、工程机械等工业设备展示。采用响应式布局技术,确保各类设备参数和图片在不同终端上都能清晰展示。查看源码
  • (响应式)企业管理人力资源服务类pbootcms模板源码下载为人力资源服务及企业管理设计的响应式网站模板,基于PbootCMS内核开发。通过宽屏布局优化岗位展示效果,简洁界面聚焦人才服务核心业务,自适应技术确保在PC端与手机端查看源码
  • 自适应手机版五金机械阀门设备通用行业网站模板该PbootCMS内核开发的网站模板适用于阀门设备、五金机械类企业,通过更换文字图片也可快速适配其他工业领域。采用响应式设计,确保PC端与手机端数据同步,操作便捷,并附带测试数据。查看源码
  • (自适应)帝国cms7.5模板新闻资讯门户带会员中心基于帝国CMS7.5内核开发的HTML5响应式模板,为新闻机构、媒体门户及资讯聚合平台设计。通过模块化布局实现图文混排查看源码
  • (自适应)宽屏大气的净水器智能电子设备网站pbootcms源码下载本模板基于PbootCMS内核开发,为净水器设备、智能电子设备企业量身打造,采用响应式设计技术,可快速构建专业级企业官网。通过本模板可高效展示产品技术参数、解决方案及企业服务优势。查看源码
  • (自适应)WordPress主题SEO自媒体博客资讯模板RabbitV2.0Rabbit v2.0主题专注于网站搜索引擎优化需求,为博客、自媒体及资讯类网站提供专业的SEO技术解决方案。该主题从架构设计到功能实现均围绕搜索引擎优化理念展开。查看源码
分享笔记 (共有 篇笔记)
验证码:

本栏推荐