Sphinx的一个应用实例

我把我blog的数据(中英文混合)导出, 作为数据来源, 来说明sphinx的使用.

准备数据源

导入数据:

mysql -u root -p test < wp_posts.sql

配置Sphinx

配置data source:

1. source blog
2. {
3.         type            =       mysql
4.         sql_host        =       localhost
5.         sql_user        =       root
6.         sql_pass        =       xxxx
7.         sql_db          =       test
8.         sql_port        =       3306
9.         sql_query       =       /
10.                 SELECT ID, post_author, UNIX_TIMESTAMP(post_date) as date_added, post_content from wp_posts
11.         sql_attr_uint   =       post_author
12.         sql_attr_timestamp =    date_added
13.         sql_query_info  =       SELECT * FROM wp_posts where ID=$id
16. }

配置index:

1. index blog
2. {
3.         source                                  = blog
4.         path                                    = /usr/local/sphinx/var/data/blog
5.         docinfo                                 = extern
6.         charset_type                    = zh_cn.utf-8
7.         charset_dictpath                = /usr/local/sphinx/dict
8. }

注意: 字典的目录要加上.

索引

./bin/indexer --config ./etc/sphinx.conf blog

如果没有错误, 会看到:

Coreseek Full Text Server 3.0(beta)

 Copyright (c) 2006-2008 coreseek.com

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

indexing index 'blog'...

collected 165 docs, 0.2 MB

sorted 0.0 Mhits, 100.0% done

total 165 docs, 164834 bytes

total 0.099 sec, 1670067.52 bytes/sec, 1671.75 docs/sec

测试搜索

./bin/search -c ./etc/sphinx.conf -i blog 苹果

输出:

Coreseek Full Text Server 3.0(beta)

 Copyright (c) 2006-2008 coreseek.com

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

0x815e1f8index 'blog': query '苹果 ': returned 1 matches of 1 total in 0.005 sec

displaying matches:

1. document=140, weight=1, post_author=1, date_added=Fri Nov 30 11:22:02 2007

ID=140

post_author=1

post_date=2007-11-30 11:22:02

post_date_gmt=2007-11-30 03:22:02

post_content=This is the text of the Commencement address by Steve Jobs, CEO of Apple Computer and of Pixar Animation Studios, delivered on June 12, 2005.

....

{中间的blog贴内容省略}

....

post_title=You've got to find what you love

post_category=0

post_excerpt=

post_status=publish

comment_status=open

ping_status=open

post_password=

post_name=youve-got-to-find-what-you-love

to_ping=

pinged=

post_modified=2008-06-20 11:58:53

post_modified_gmt=2008-06-20 03:58:53

post_content_filtered=

post_parent=0

guid=http://blog.funcat.cn/?p=164

menu_order=0

post_type=post

post_mime_type=

comment_count=0

words:

1. '苹果': 1 documents, 10 hits

启动searchd

./bin/searchd

Coreseek Full Text Server 3.0(beta)

 Copyright (c) 2006-2008 coreseek.com

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

使用调用API

1. #!/usr/bin/python
2. # -*- coding:utf-8 -*-
4. import sys
5. if sys.getdefaultencoding() != 'utf-8': 
6.     reload(sys) 
7.     sys.setdefaultencoding('utf-8')
9. import web
10. from web.contrib.template import render_mako
12. import MySQLdb
13. from MySQLdb import *
15. from sphinxapi import *
17. urls = (
18.     '/', 'index',
19. )
21. render = render_mako(
22.         directories=['templates'],
23.         input_encoding='utf-8',
24.         output_encoding='utf-8',
25.         )
27. app = web.application(urls, globals())
28. con = MySQLdb.Connect(host="localhost", port=3306, user="root", passwd="xixihaha", db="blogdata")
30. class index:
31.     def GET(self):
32.         r_info = ''
33.         info = ''
34.         s_result = ''
35.         return render.index(r_info=r_info, e_info=info, s_result=s_result)
37.     def POST(self):
38.         i = web.input()
39.         if i.keyword == '':
40.             raise web.seeohter('/')
41.         
42.         e_info = ''
43.         r_info = ''
44.         s_result = ''
45.         
46.         q = i.keyword
47.         
48.         cl = SphinxClient()
49.         cl.SetServer ( 'localhost', 3312 )
50.         res = cl.Query ( q, 'blog' )
51.         if not res:
52.             e_info = 'query failed: %s' % cl.GetLastError()
53.             
54.         if cl.GetLastWarning():
55.             e_info = 'WARNING: %s/n' % cl.GetLastWarning()
57.         if res.has_key('words'):
58.             for info in res['words']:
59.                 r_info += '/t/'%s/' found %d times in %d documents<br/>' % (info['word'], info['hits'], info['docs'])
60.                 
61.         if res.has_key('matches'):
62.             n = 1
63.             s_result = '/nMatches:<br/>'
64.             
65.             import time
66.             
67.             print res['matches']
68.             
69.             for match in res['matches']:
70.                 attrsdump = ''
71.                 for attr in res['attrs']:
72.                     attrname = attr[0]
73.                     attrtype = attr[1]
74.                     value = match['attrs'][attrname]
75.                     if attrtype==SPH_ATTR_TIMESTAMP:
76.                         value = time.strftime ( '%Y-%m-%d %H:%M:%S', time.localtime(value) )
77.                     attrsdump = '%s, %s=%s' % ( attrsdump, attrname, value )
78.         
79.                     s_result += '%d. doc_id=%s, weight=%d%s<br/>' % (n, match['id'], match['weight'], attrsdump)
80.                 n += 1
81.                 Cursor = con.cursor()
82.                 
83.                 Cursor.execute('select post_content from wp_posts where id = %s' % match['id'])
84.                 re = Cursor.fetchall()
85.                 s_result += re[0][0]
86.                 s_result += '<hr/>'
87.         
88.         return render.index(r_info=r_info, e_info=info, s_result=s_result)
89.         
90. if __name__ == '__main__':
91.     app.run()