csft-3.1搜索问题

我在网上找了一个安装csft-3.1的过程,步骤一模一样
只是在/etc/csft中找不到csfs.conf,所以复制一份当前目录的sphinx.conf改名为csfs.conf
而且也把mysql的编码改过来了
还以防万一删除数据库又重新建立了数据库,创建时加了句character set utf-8
但是英文可以搜索出来中文却不行
为什么会这样呢?

以下是我在网上找的安装过程

安装过程：
wget http://www.coreseek.cn/uploads/csft/3.1/CentOS5/mmseg-3.1-1.i386.rpm
wget http://www.coreseek.cn/uploads/csft/3.1/CentOS5/csft-3.1-1.1.i386.rpm
安装csft-3.1-1.1.i386.rpm的时候会提示需要个postgresql的动态库的依赖，需要安装
yum install -y postgresql-libs.i386
rpm -Uvh csft-3.1-1.1.i386.rpm mmseg-3.1-1.i386.rpm
（另外需要下载mmseg的源代码包)
把安装后的/etc/csft/example.sql倒入数据库
mysql < example.sql
wget http://www.coreseek.cn/uploads/csft/3.1/Source/mmseg-3.1.tar.gz
tar zxvf mmseg-3.1.tar.gz
cd mmseg-3.1/data
mmseg -u unigram.txt
把生成的unigram.txt.uni 改名并拷贝到相应位置
mv unigram.txt.uni /var/data/dict/uni.lib
cd /etc/csft
vi csfs.conf
编辑索引定义添加
charset_type        = zh_cn.utf-8
charset_dictpath    = /var/data/dict
保存建立索引
# csft-indexer –all (对所有索引定义建立索引）
Coreseek Full Text Server 3.1
Copyright (c) 2006-2008 coreseek.com
using config file ‘./csft.conf’…
indexing index ‘test1′…
iniparser: cannot open /var/data/dict/mmseg.ini
1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        1,
pt:1, 1;        3, 6,
pt:3, 24;       pt:6, 0;        3, 6,
pt:3, 47279;    pt:6, 0;        3, 6,
pt:3, 61;       pt:6, 0;        3,
pt:3, 13411;    pt:3, 30471;    pt:6, 1;        pt:3, 30471;    pt:3, 24;       pt:6, 1;        pt:3, 24;       pt:6, 0;        1,
3, 6,
pt:1, 1;        pt:3, 24;       pt:6, 1;        pt:3, 24;       pt:6, 0;        3,
pt:3, 14538;    pt:3, 298;      3, 6,
pt:3, 24;       pt:6, 0;        3, 6,
pt:3, 24;       pt:3, 154;      pt:1, 1;        pt:6, 0;        pt:1, 1;        pt:3, 13411;    pt:1, 1;        pt:3, 13411;    pt:3, 13411;    pt:3, 30471;    3,
pt:3, 990;      pt:3, 1448;     pt:6, 1;        pt:3, 1448;     pt:6, 0;        collected 8 docs, 0.0 MB
sorted 0.0 Mhits, 100.0% done
total 8 docs, 269 bytes
total 0.030 sec, 8825.17 bytes/sec, 262.46 docs/sec
然后测试中文分词
# csft-search 测试
Coreseek Full Text Server 3.1
Copyright (c) 2006-2008 coreseek.com
using config file ‘/etc/csft/csft.conf’…
iniparser: cannot open /var/data/dict/mmseg.ini
3, 6,
pt:3, 24;       pt:3, 154;      pt:6, 0;        pt:1, 1;        index ‘test1′: query ‘测试 ‘: returned 4 matches of 4 total in 0.014 sec
displaying matches:
1. document=7, weight=2, group_id=3, date_added=Tue Jan 12 15:37:42 2010
id=7
group_id=3
group_id2=11
date_added=2010-01-12 15:37:42
title=12 测试
content=大册，测试
2. document=5, weight=1, group_id=3, date_added=Tue Jan 12 13:38:04 2010
id=5
group_id=3
group_id2=9
date_added=2010-01-12 13:38:04
title=测试
content=一些
3. document=6, weight=1, group_id=3, date_added=Tue Jan 12 15:26:01 2010
id=6
group_id=3
group_id2=10
date_added=2010-01-12 15:26:01
title=标题
content=我的测试
4. document=8, weight=1, group_id=3, date_added=Tue Jan 12 15:42:25 2010
id=8
group_id=3
group_id2=12
date_added=2010-01-12 15:42:25
title=测试 我的
content=先吃饭
words:
1. ‘测试’: 4 documents, 5 hits

已经解决了...
是csfs.conf配置文件中没把sql_query_pre    = SET NAMES utf8注释去掉
但是现在出现一个问题
搜索结果中中文是以问号显示


[root@localhost csft]# csft-search 测试
Coreseek Full Text Server 3.1
Copyright (c) 2006-2008 coreseek.com
using config file '/etc/csft/csft.conf'...
3, 6,
pt:3, 24;       pt:3, 154;      pt:6, 0;        pt:1, 1;        index 'test1': query '测试 ': returned 3 matches of 3 total in 0.007 sec

displaying matches:
1. document=5, weight=2, group_id=1, date_added=Thu May 13 10:21:28 2010
        id=5
        group_id=1
        group_id2=5
        date_added=2010-05-13 10:21:28
        title=???
        content=????????????????????
2. document=6, weight=2, group_id=1, date_added=Thu May 13 10:21:28 2010
        id=6
        group_id=1
        group_id2=6
        date_added=2010-05-13 10:21:28
        title=???
        content=?????????
3. document=8, weight=1, group_id=2, date_added=Thu May 13 10:21:28 2010
        id=8
        group_id=2
        group_id2=8
        date_added=2010-05-13 10:21:28
        title=????
        content=??????

words:
1. '测试': 3 documents, 5 hits


终端编码：utf8

mysql：
ysql> show variables like 'character_set_%';
+--------------------------+----------------------------+
| Variable_name            | Value                      |
+--------------------------+----------------------------+
| character_set_client     | utf8                       |
| character_set_connection | utf8                       |
| character_set_database   | utf8                       |
| character_set_filesystem | binary                     |
| character_set_results    | utf8                       |
| character_set_server     | utf8                       |
| character_set_system     | utf8                       |
| character_sets_dir       | /usr/share/mysql/charsets/ |
+--------------------------+----------------------------+
8 rows in set (0.00 sec)

mysql> show variables like 'collation_%';
+----------------------+-----------------+
| Variable_name        | Value           |
+----------------------+-----------------+
| collation_connection | utf8_general_ci |
| collation_database   | utf8_general_ci |
| collation_server     | utf8_general_ci |
+----------------------+-----------------+
3 rows in set (0.00 sec)


csft.conf：
source src1
{
    type = mysql

    sql_host = localhost
    sql_user = test
    sql_pass =
    sql_db = test
    sql_port = 3306

    sql_query_pre = SET NAMES utf8

    sql_query = \
        SELECT id, group_id, UNIX_TIMESTAMP(date_added) AS date_added, title, content \
        FROM documents

    sql_attr_uint            = group_id

    sql_attr_timestamp        = date_added

    sql_ranged_throttle    = 0

    sql_query_info        = SELECT * FROM documents WHERE id=$id
}

source src1throttled : src1
{
    sql_ranged_throttle = 100
}

index test1
{
    source = src1
    path = /var/data/test1
    docinfo = extern
    mlock = 0
    morphology = none
    min_word_len = 1

    charset_type = zh_cn.utf-8
    charset_dictpath = /var/data/dict
    html_strip = 0
}

indexer
{
    mem_limit = 32M
}

searchd
{
    log = /var/log/searchd.log
    query_log = /var/log/query.log
    read_timeout = 5
    client_timeout = 300
    max_children = 30
    pid_file = /var/log/searchd.pid
    max_matches = 1000
    seamless_rotate = 1
    preopen_indexes = 0
    unlink_old = 1
    mva_updates_pool = 1M
    max_packet_size = 8M
    max_filters     = 256
    max_filter_values = 4096
}

请注意数据库原始数据的编码，以及conf的设置编码，query_pre的编码请求设置，以及命令行界面本身的编码都需要一致。

我在2楼贴上已经写的很清楚了
所有的都配置好了
而且我把原来的数据库删了后重新用create database test character set utf-8;建立了数据库
但是在终端显示的中文依然是问号
在MYSQL中查询表内容是显示正常的

那你的mysql的客户端编码是什么？

以下是我机器配置的信息
还需要配置什么其他部分吗?

mysql：
mysql> show variables like 'character_set_%';
+--------------------------+----------------------------+
| Variable_name | Value |
+--------------------------+----------------------------+
| character_set_client | utf8 |
| character_set_connection | utf8 |
| character_set_database | utf8 |
| character_set_filesystem | binary |
| character_set_results | utf8 |
| character_set_server | utf8 |
| character_set_system | utf8 |
| character_sets_dir | /usr/share/mysql/charsets/ |
+--------------------------+----------------------------+
8 rows in set (0.00 sec)

mysql> show variables like 'collation_%';
+----------------------+-----------------+
| Variable_name | Value |
+----------------------+-----------------+
| collation_connection | utf8_general_ci |
| collation_database | utf8_general_ci |
| collation_server | utf8_general_ci |
+----------------------+-----------------+
3 rows in set (0.00 sec)

csft.conf：
sql_query_pre = SET NAMES utf8
charset_type = zh_cn.utf-8

终端：utf8

系统编码：
LANG=zh_CN.UTF-8
LC_CTYPE="zh_CN.UTF-8"
LC_NUMERIC="zh_CN.UTF-8"
LC_TIME="zh_CN.UTF-8"
LC_COLLATE="zh_CN.UTF-8"
LC_MONETARY="zh_CN.UTF-8"
LC_MESSAGES="zh_CN.UTF-8"
LC_PAPER="zh_CN.UTF-8"
LC_NAME="zh_CN.UTF-8"
LC_ADDRESS="zh_CN.UTF-8"
LC_TELEPHONE="zh_CN.UTF-8"
LC_MEASUREMENT="zh_CN.UTF-8"
LC_IDENTIFICATION="zh_CN.UTF-8"
LC_ALL=

sql_query_pre = SET NAMES utf8

如果你的数据库是正确的编码的话，那么上面这一句，是必要的。