08 七 10 Sphinx使用入门

搜索三大要素：数据来源、预处理、查询。 
在Sphinx+MySQL的架构中，MySQL主要提供了数据来源和查询接口，真正进行全文索引建立和查询的是Sphinx。 
MySQL里面存放真正的数据；Sphinx从MySQL中获取数据建立全文索引；应用程序使用相应的api与Sphinx交互以获得真正的数据（此处的api包含SQL接口、php接口，以及其他一些编程语言能够调用的接口）。 

假设test库内有表test，结构如下： 
CREATE TABLE `test` ( 
`id` bigint(20) unsigned NOT NULL AUTO_INCREMENT, 
`content` varchar(20000) DEFAULT NULL, 
`gmt_create` datetime DEFAULT NULL, 
`v` char(100) DEFAULT NULL, 
PRIMARY KEY (`id`) 
) ENGINE=InnoDB DEFAULT CHARSET=utf8 
我们需要在content字段上面做全文索引。 

根据以上信息配置Sphinx是一件相对容易的事情。在只涉及到单数据来源（MySQL）的情况下，我们只需要关注下面这几个参数： 
source src1 
{ 
# data source type. mandatory, no default value 
# known types are mysql, pgsql, mssql, xmlpipe, xmlpipe2, odbc 
type = mysql 

##################################################################### 
## SQL settings (for ‘mysql’ and ‘pgsql’ types) 
##################################################################### 

# some straightforward parameters for SQL source types 
sql_host = 192.168.*.* 
sql_user = realzyy 
sql_pass = realzyy 
sql_db = test 
sql_port = 3306 # optional, default is 3306 

# pre-query, executed before the main fetch query 
# multi-value, optional, default is empty list of queries 
# 
sql_query_pre = SET NAMES utf8 
# sql_query_pre = SET SESSION query_cache_type=OFF 

# range query setup, query that must return min and max ID values 
# optional, default is empty 
# 
# sql_query will need to reference $start and $end boundaries 
# if using ranged query: 
# 
sql_query = \ 
SELECT id, gmt_create, content,v \ 
FROM test \ 
WHERE id>=$start AND id<=$end 

sql_query_range = SELECT MIN(id),MAX(id) FROM test 

# range query step 
# optional, default is 1024 
# 
sql_range_step = 128 

# UNIX timestamp attribute declaration 
# multi-value (an arbitrary number of attributes is allowed), optional 
# similar to integer, but can also be used in date functions 
# 
# sql_attr_timestamp = posted_ts 
# sql_attr_timestamp = last_edited_ts 
sql_attr_timestamp = gmt_create 

# string ordinal attribute declaration 
# multi-value (an arbitrary number of attributes is allowed), optional 
# sorts strings (bytewise), and stores their indexes in the sorted list 
# sorting by this attr is equivalent to sorting by the original strings 
# 
sql_attr_str2ordinal = v 

# ranged query throttling, in milliseconds 
# optional, default is 0 which means no delay 
# enforces given delay before each query step 
sql_ranged_throttle = 0 

# document info query, ONLY for CLI search (ie. testing and debugging) 
# optional, default is empty 
# must contain $id macro and must fetch the document by that id 
sql_query_info = SELECT * FROM test WHERE id=$id 
} 

因为这个表的字符集采用了utf8，所以还需要修改一下： 
index test1 
{ 
source = src1 
# document source(s) to index 
# multi-value, mandatory 
# document IDs must be globally unique across all sources 
source = src1 

# index files path and file name, without extension 
# mandatory, path must be writable, extensions will be auto-appended 
path = /u01/sphinx/var/data/index_for_test 

# document attribute values (docinfo) storage mode 
# optional, default is ‘extern’ 
# known values are ‘none’, ‘extern’ and ‘inline’ 
docinfo = extern 

# memory locking for cached data (.spa and .spi), to prevent swapping 
# optional, default is 0 (do not mlock) 
# requires searchd to be run from root 
mlock = 0 

# a list of morphology preprocessors to apply 
# optional, default is empty 
# 
# builtin preprocessors are ‘none’, ‘stem_en’, ‘stem_ru’, ‘stem_enru’, 
# ‘soundex’, and ‘metaphone’; additional preprocessors available from 
# libstemmer are ‘libstemmer_XXX’, where XXX is algorithm code 
# (see libstemmer_c/libstemmer/modules.txt) 
# 
# morphology = stem_en, stem_ru, soundex 
# morphology = libstemmer_german 
# morphology = libstemmer_sv 
morphology = none 
# minimum indexed word length 
# default is 1 (index everything) 
min_word_len = 1 

# charset encoding type 
# optional, default is ‘sbcs’ 
# known types are ‘sbcs’ (Single Byte CharSet) and ‘utf-8′ 
charset_type = utf-8 

# charset definition and case folding rules “table” 
# optional, default value depends on charset_type 
# 
# defaults are configured to include English and Russian characters only 
# you need to change the table to include additional ones 
# this behavior MAY change in future versions 
# 
# ‘sbcs’ default value is 
# charset_table = 0..9, A..Z->a..z, _, a..z, U+A8->U+B8, U+B8, U+C0..U+DF->U+E0..U+FF, U+E0..U+FF 
# 
# ‘utf-8′ default value is 
charset_table = 0..9, A..Z->a..z, _, a..z, U+410..U+42F->U+430..U+44F, U+430..U+44F 
} 
执行一下indexer创建全文索引；再执行searchd打开Sphinx的监听端口以便接收请求；最后可以用search去检测一下战果~ 
#indexer –all 
#searchd 
#search *****