Installation from source
choose a directory to install all softwares, as an example we take
- You can know download all required softwares with a download.sh script which is inside the distribution!
The script use curl to download.
- Install pre-requisite softwares, if you don't have them:
Follow the procedure from sqlite web site,
we use 3.2.7 version.
If you wan't to minimize the memory consumption at indexation time, you can
install this hash map program.
Follow the procedure from
sparesehash web site,
we use 0.3 version. If you use 0.2 version, you may have to add
a make CXXFLAGS+='-Isrc' in order to have it to build.
This is a BSD implementation of the readline
library. If you don't install it, you just loose the history of
command for ft3_searcher.
You can download it from thrysoee.
You need expat to build the demo. It should already be
available on your site.
NOT USED IN 0.3, WILL BE IN 0.4
Follow the procedure from ICU download web site, we
use 3.2 version.
- grad the tarball from sourceforge.net.
- decompess it somewhere:
tar zxvf ft3-0.3.0.tar.gz
and change directory via
- configure the software:
sh ./configure --prefix=$HOME/dist
sh ./configure --help
gives a full list of options. Note that if you want to install in another directory
with the prefix option, you have to add some flags at configure time:
CPPFLAGS='-I/where/I/install/include' LDFLAGS='-L/where/I/install/lib' sh ./configure --prefix=/where/I/install
- compile it with a
and install it with a:
Note: you get some warnings when you compile ft3_language which you can ignore.
Each version of FT3 should come with some binaries for platform I have access
to, currently cygwin, WinXP and Linux 32bits (Ubuntu). You just grab
a tarball for your OS, untar it (or unzip it) and you got the 2 binaries
build in the previous section. You still need the source tarball to get
the documentation and SQL schemas.