---

layout: post

title: "Debian/Linux下Sphinx-for-chinese (中文全文搜索)的安装"

---

Sphinx是一个基于SQL的全文检索引擎，但对中文用户来说一个致命的缺陷是不支持中文。后来在网上发现了一个基于 Sphinx 的支持切词的全文搜索引擎 <a href="http://code.google.com/p/sphinx-for-chinese/" target="_blank">sphinx-for-chinese</a>。下载下来安装使用后发现很好用，下面介绍一下具体的安装过程。

<span id="more-720"></span>

<ol>

<li><strong>下载所需的安装包</strong>

sphinx-for-chinese-0.9.9-r2117.tar.gz

xdict_1.1.tar.gz

下载地址：<a href="http://code.google.com/p/sphinx-for-chinese/downloads/list" target="_blank">http://code.google.com/p/sphinx-for-chinese/downloads/list</a></li>

<li>安装 sphinx-for-chinese

<pre>$ tar zxvf sphinx-for-chinese-0.9.9-r2117.tar.gz

$ cd sphinx-for-chinese-0.9.9-r2117

$ ./configure --prefix=/usr/local/sphinx

$ make

$ sudo make install</pre>

</li>

<li>创建test数据库，并创建sphinx用户

<pre>mysql&gt; create database test;

mysql&gt;create user 'sphinx'@'localhost' identified by 'sphinx';

mysql&gt;grant all privileges on test.* to 'sphinx'@'localhost';</pre>

</li>

<li>指定sphinx配置文件

<pre>$ cd /usr/local/sphinx/etc

$ sudo cp sphinx.conf.dist sphinx.conf</pre>

</li>

<li>编辑配置文件

<pre>sql_host        = localhost

<strong>sql_user        = sphinx

sql_pass        = sphinx</strong>

sql_db          = test

sql_port        = 3306  # optional, default is 3306</pre>

说明：加粗部分是修改的内容</li>

</ol>

到这里为止，sphinx已经可以使用了，但还不能支持中文切词,以下是加入中文切词的步骤

<ol>

<li>解压字典文件 xdict_1.1.tar.gz

<pre>$ tar zxvf xdict_1.1.tar.gz</pre>

</li>

<li>借助先前安装的 mkdict 工具生成字典

<pre>$ /usr/local/sphinx/bin/mkdict xdict.txt xdict</pre>

</li>

<li>将字典 xdict 拷贝到 /usr/local/sphinx/etc目录下</li>

<li>配置中文切词

打开 sphinx.conf文件，找到 'charset_type    = sbcs' 字样，将其改为

<pre>charset_type    = utf-8

chinese_dictionary = /usr/local/sphinx/etc/xdict</pre>

</li>

</ol>

至此中文切词配置完成，下面做一个简单的测试

<ol>

<li>编辑sphinx-for-chinese自带的SQL脚本，加入中文数据

$ vi /usr/local/sphinx/etc/example.sql

<pre>REPLACE INTO test.documents ( id, group_id, group_id2, date_added, title, content ) VALUES

( 1, 1, 5, NOW(), 'test one', 'this is my test document number one. also checking search within phrases.' ),

( 2, 1, 6, NOW(), 'test two', 'this is my test document number two' ),

( 3, 2, 7, NOW(), 'another doc', 'this is another group' ),

( 4, 2, 8, NOW(), 'doc number four', 'this is to test groups' ),

<strong>( 5, 2, 8, NOW(), 'doc number five', '一个' ),

( 6, 2, 8, NOW(), 'doc number six', '我' ),

( 7, 2, 8, NOW(), 'doc number seven', '中国人' )</strong>;

</pre>

说明：加粗部分是添加的中文测试数据

</li>

<li>导入数据

<pre>$ mysql -usphinx -psphinx &lt; example.sql</pre>

</li>

<li>建立索引

<pre>$ sudo /usr/local/sphinx/bin/indexer --all</pre>

</li>

<li>检索

<pre>$ /usr/local/sphinx/bin/search 我是一个中国人

Sphinx 0.9.9-release (r2117)

Copyright (c) 2001-2009, Andrew Aksyonoff

using config file '/usr/local/sphinx/etc/sphinx.conf'...

index 'test1': query '我是一个中国人 ': returned 0 matches of 0 total in

0.000 sec

words:

1. '我': 1 documents, 1 hits

2. '是': 0 documents, 0 hits

3. '一个': 1 documents, 1 hits

4. '中国人': 1 documents, 1 hits

index 'test1stemmed': query '我是一个中国人 ': returned 0 matches of 0

total in 0.000 sec

words:

1. '我': 1 documents, 1 hits

2. '是': 0 documents, 0 hits

3. '一个': 1 documents, 1 hits

4. '中国人': 1 documents, 1 hits</pre>

</li>

</ol>

至此，sphinx-for-chinese已经成功安装并顺利通过测试