INSTALLATION

REQUIREMENTS

CHASEN

In order to use JSplitter, you first have to have the ChaSen library installed on your system. ChaSen is available at:

http://cl.aist-nara.ac.jp/lab/nlt/chasen/

Follow the instructions in the ChaSen README to install. Note that make install does not install the libchasen library, but does install dictionaries and settings files; I suggest copying libchasen.a (which will appear in the lib/ directory after building) to somewhere in your library path... a good place is /usr/local/lib. The .h files from ChaSen's lib/ also need to be installed somewhere where Setup can find them (I put them in /usr/local/lib; if you put them somewhere else you might need to add -I/ to Setup).

KCONV

Kconv is a package for doing Japanese character encoding detection and conversion, used by JSplitter. Kconv is available at:

http://hito.eccosys.com/~mak/kconv/

Choose either the Python or C++ version and install the Kconv module somewhere in your PYTHONPATH.

ExtensionClass

ExtensionClass is needed for the new sub-classable Splitter used in this distribution. ExtensionClass is included with any Zope source distribution, or you can install it separately.

BUILD

Follow the usual drill to build the extension modules::

# make -f Makefile.pre.in boot # make

Move the resulting shared libraries, (probably) Splitter.so and chalibmodule.so, up one directory after they build successfully.

Here is a brief description of the modules and what they are for:

• Splitter.c - This is a EUC (Codeset 1 & 2) safe, ExtensionClass version of the standard Splitter.
• chalibmodule.c - This is an extension module to expose the barest minimum of functionality from chalib to be able to use its parsing and analysis facilities.