csft安装过程中出错的问题及解决办法（目前仍然无法成功进行对中文的处理）...

1 . 安装mmesg

 wget http://www.coreseek.cn/uploads/csft/3.1/Source/mmseg-3.1.tar.gz
 # ./configure --prefix&＃61;/usr/local/mmseg && make && make install   

2 . 安装PYTHON2.5
  wget http://www.python.org/ftp/python/2.5/Python-2.5.tar.bz2
  ./configure 
3 . 设置环境变量
# cd -
# vim .profile

增加两行
export CPPFLAGS&＃61;-I/usr/local/include/python2.5
export LDFLAGS&＃61;-lpython2.5

4. 安装csft
   wget http://www.coreseek.cn/uploads/csft/3.1/Source/csft-3.1.tar.gzhttp://www.coreseek.cn/uploads/csft/3.1/Source/csft-3.1.tar.gz

./configure --prefix&＃61;/usr/local/webserver/csft --with-mysql&＃61;/usr/local/webserver/mysql/ --with-mmseg-includes&＃61;/usr/local/mmseg/include/mmseg --with-mmseg-libs&＃61;/usr/local/mmseg/lib/ --without-python



You have new mail in /var/spool/mail/zz




$make
Making all in src
make[1]: Entering directory &＃96;/home/zyf/zyfwork/csft3.1b3/src&＃39;
if test -d ../.svn; then svn info .. --xml | perl svnxrev.pl; fi;
make  all-am
make[2]: Entering directory &＃96;/home/zyf/zyfwork/csft3.1b3/src&＃39;
if g&＃43;&＃43; -DHAVE_CONFIG_H -I. -I. -I../config  -DSYSCONFDIR&＃61;"\"/usr/local/webserver/csft/etc\"" -I/usr/local/include -pthread -I/usr/local/webserver/mysql/include/mysql -DUNIV_LINUX  -I/usr/local/include/mmseg   -Wall -g -D_FILE_OFFSET_BITS&＃61;64 -O3 -DNDEBUG -MT tokenizer_zhcn.o -MD -MP -MF ".deps/tokenizer_zhcn.Tpo" -c -o tokenizer_zhcn.o tokenizer_zhcn.cpp; \
then mv -f ".deps/tokenizer_zhcn.Tpo" ".deps/tokenizer_zhcn.Po"; else rm -f ".deps/tokenizer_zhcn.Tpo"; exit 1; fi
In file included from /usr/local/include/mmseg/Segmenter.h:38,
                 from /usr/local/include/mmseg/SegmenterManager.h:32,
                 from tokenizer_zhcn.cpp:1:
/usr/local/include/mmseg/mmthunk.h: In member function &＃96;u2 css::ChunkQueue::getToken()&＃39;:
/usr/local/include/mmseg/mmthunk.h:140: warning: comparison between signed and unsigned integer expressions
/usr/local/include/mmseg/mmthunk.h:157: warning: converting of negative value &＃96;-0x000000001&＃39; to &＃96;size_t&＃39;
/usr/local/include/mmseg/mmthunk.h:158: warning: comparison between signed and unsigned integer expressions
tokenizer_zhcn.cpp: In member function &＃96;css::Segmenter* CSphTokenizer_zh_CN_UTF8_Private::GetSegmenter(const char*)&＃39;:
tokenizer_zhcn.cpp:54: error: no matching function for call to &＃96;css::SegmenterManager::getSegmenter(bool)&＃39;
/usr/local/include/mmseg/SegmenterManager.h:46: note: candidates are: css::Segmenter* css::SegmenterManager::getSegmenter()
tokenizer_zhcn.cpp: In member function &＃96;virtual bool CSphTokenizer_zh_CN_UTF8::IsSentenceEnd()&＃39;:
tokenizer_zhcn.cpp:155: error: &＃39;class css::Segmenter&＃39; has no member named &＃39;isSentenceEnd&＃39;
tokenizer_zhcn.cpp: In member function &＃96;virtual bool CSphTokenizer_zh_CN_UTF8::IsKeyWord(BYTE*, int)&＃39;:
tokenizer_zhcn.cpp:161: error: &＃39;class css::Segmenter&＃39; has no member named &＃39;isKeyWord&＃39;
tokenizer_zhcn.cpp: In member function &＃96;virtual int CSphTokenizer_zh_CN_UTF8::GetWordWeight(BYTE*, int) const&＃39;:
tokenizer_zhcn.cpp:167: error: &＃39;class css::Segmenter&＃39; has no member named &＃39;getWordWeight&＃39;
tokenizer_zhcn.cpp: In member function &＃96;virtual BYTE* CSphTokenizer_zh_CN_UTF8::GetToken(int)&＃39;:
tokenizer_zhcn.cpp:180: error: no matching function for call to &＃96;css::Segmenter::peekToken(u2&, u2&, int&)&＃39;
/usr/local/include/mmseg/Segmenter.h:110: note: candidates are: const u1* css::Segmenter::peekToken(u2&, u2&)
tokenizer_zhcn.cpp:181: error: no matching function for call to &＃96;css::Segmenter::popToken(u2&, int&)&＃39;
/usr/local/include/mmseg/Segmenter.h:111: note: candidates are: void css::Segmenter::popToken(u2)
make[2]: *** [tokenizer_zhcn.o] Error 1
make[2]: Leaving directory &＃96;/home/zyf/zyfwork/csft3.1b3/src&＃39;
make[1]: *** [all] Error 2
make[1]: Leaving directory &＃96;/home/zyf/zyfwork/csft3.1b3/src&＃39;
make: *** [all-recursive] Error 1



/usr/local/sphinx/src/sphinx.cpp:15557: undefined reference to &＃96;libiconv_open&＃39;
libsphinx.a(sphinx.o)(.text&＃43;0x53a01):/usr/local/sphinx/src/sphinx.cpp:15575: undefined
reference to &＃96;libiconv&＃39;
libsphinx.a(sphinx.o)(.text&＃43;0x53a28):/usr/local/sphinx/src/sphinx.cpp:15581: undefined
reference to &＃96;libiconv_close&＃39;
collect2: ld returned 1 exit status
make[2]: * [indexer] Error 1
make[2]: Leaving directory &＃96;/usr/local/sphinx/src&＃39;
make[1]: * [all] Error 2
make[1]: Leaving directory &＃96;/usr/local/sphinx/src&＃39;
make: * [all-recursive] Error 1

官网解决办法&＃xff1a;

In the meantime I&＃39;ve change the configuration file and set
#define USE_LIBICONV 0 in line 8179.
修改configure 文件把 #define USE_LIBICONV 0 最后的数值由1改为0
重新编译。

我使用的参数如下&＃xff1a;
./configure --prefix&＃61;/usr/local/coreseek --with-mysql&＃61;/opt/lampp --with-mysql-includes&＃61;/opt/lampp/include/mysql/ --with-mysql-libs&＃61;/opt/lampp/lib/mysql/ --with-mmseg-includes&＃61;/usr/local/mmseg/include/mmseg --with-mmseg-libs&＃61;/usr/local/mmseg/lib/


编译错误一
make[2]: *** [indexer] Error 1
make[2]: Leaving directory &＃96;/www/tmp/csft-3.1/src&＃39;
make[1]: *** [all] Error 2
make[1]: Leaving directory &＃96;/www/tmp/csft-3.1/src&＃39;
make: *** [all-recursive] Error 1

解决办法

vi ./src/sphinx.cpp

注释以下代码

#case TOKENIZER_ZHCN_GBK:
#pTokenizer &＃61; sphCreateGBKChineseTokenizer
#(tSettings.m_sDictPath.cstr(), tSettings.m_nBest); break;
然后重新编译
make clean

make

make install


错误二&＃xff0c;索引出错
[root&＃64;localhost csft-3.1]# /usr/local/coreseek/bin/indexer --config /usr/local/coreseek/etc/sphinx.conf --all
/usr/local/coreseek/bin/indexer: error while loading shared libraries: libmysqlclient.so.16: cannot open shared object file: No such file or directory

解决办法

ln -s /opt/lampp/lib/mysql/libmysqlclient.so.16 /usr/lib/libmysqlclient.so.16

错误三&＃xff0c;索引不生成

解决办法&＃xff0c;原来是手误&＃xff0c;把/usr/local/coreseek/bin/indexer --config /usr/local/coreseek/etc/sphinx.conf --all
写成
/usr/local/coreseek/bin/indexer --config /usr/local/coreseek/etc/sphinx.conf -all

错误四&＃xff0c;服务启动报错

[root&＃64;localhost coreseek]# ./bin/searchd --config /usr/local/coreseek/etc/sphinx.conf
Coreseek Full Text Server 3.1
Copyright (c) 2006-2008 coreseek.com
using config file &＃39;/usr/local/coreseek/etc/sphinx.conf&＃39;...
listening on all interfaces, port&＃61;3312
iniparser: cannot open /usr/local/coreseek/data/dict/mmseg.ini

解决办法

vi /usr/local/coreseek/data/dict/mmseg.ini

输入以下

[mmseg]
merge_number_and_ascii&＃61;1;
number_and_ascii_joint&＃61;-;
compress_space&＃61;0;
seperate_number_ascii&＃61;1;

以上解释如下

/*
merge_number_and_ascii: 字母和数字连续出现是非切分
number_and_ascii_joint:连接数字和字母可用的符号&＃xff0c;如&＃39;-&＃39; &＃39;.&＃39; 等
compress_space&＃xff1a;暂时无效
seperate_number_ascii&＃xff1a;是否拆分数字&＃xff0c;如 1988 -> 1/x 9/x 8/x 8/x

*/