Denbun POP version Manual
User's Manual | Administrator's Manual | Installation Guide | Initial Settings Guide
Back to Menu
 

Installing and configuring database search engine (MeCab and PGMeCab)

MeCab is an open source morphological analysis engine developed through the project " 京都大学情報学研究科-日本電信電話株式会社コミュニケーション科学基礎研究所共同研究ユニットプロジェクト".
Please refer the MeCab web site (http://mecab.sourceforge.net/) for more information about.

Please refer the system requirements for which version of MeCab that Denbun supports.
1. Expanding the package
Expand the file you transferred by FTP. In this example, it assumes that the file exists in "/home/guest" directory.


After the expansion, a directory named "mecab-0.96" is created in "/usr/local/src".
2. Install the package
To configure the package, use configure scripts.
Specify UTF-8 as a character set.
Move to the directory you expand the package and run the configure script.


Make the package.
User GNU make to make the package.



Now, it is ready to install the package. This step requires the super user (root).



After the checking the installation, use gmake clean command to cleaning. This command deletes unnecessary files and makes free space.

3. Create a link of the library
Since MeCab library is installed in "/usr/local/lib", create a symbolic link in "/usr/lib".

 

4. Install a dictionary for MeCab

5. Expand the package source
Expand the file you transferred by FTP. In this example, it assumes that the file exists in "/home/guest" directory.

 



After the expansion, a directory named "mecab-ipadic-2.7.0-20070801" is created in "/usr/local/src".
6. Installing the package
To configure the package, use configure scripts.
Specify UTF-8 as a character set.
Move to the directory you expand the package and run the configure script.



Make the package.
User GNU make to make the package.



Now, it is ready to install the package. This step requires the super user (root).



After the checking the installation, use gmake clean command to cleaning. This command deletes unnecessary files and makes free space.

7. Installing pgmecab()
Since Tsearch2 creates indexes by word by word, it does not work on languages such as Japanese that do not delimit words by a space.
For Japanese documents, MeCab delimits Japanese words and index them for Tsearch2.
8. Expanding the package
Expand the file you transferred by FTP. In this example, it assumes that the file exists in "/home/guest" directory.


After the expansion, a directory named "pgmecab-1.1" is created in "/usr/local/src".

9. Install the package
Before making the package, edit "pgmecab.c" and "Makefile".
First of all, add the following three lines to "/usr/local/src/pgmecab-1.1/pgmecab.c".


Next, edit "/usr/local/src/pgmecab-1.1/Makefile".
Set MECAB_CONFIG_PATH to "/usr/local/bin/mecab-config" and set USE_PGXS parameter.


Make the package.
Since it uses the administrative command of PostgreSQL, set the environment variable to the path for the administrative command. This step requires the super user (root).


User GNU make to make the package.


Now, it is ready to install the package. This step requires the super user (root).


After the checking the installation, use gmake clean command to cleaning. This command deletes unnecessary files and makes free space.


Now, "pgmecab()" installation completed.


 
Copyright (C) 2008-2011 NEOJAPAN,Inc. All Rights Reserved.