Because this tool uses the FuXi inference engine, it requires a certain branch of RDFLib, namelay the layercake branch. Installing is nonetheless easy enough. It is recommended to make use of virtualenv to create a python environment for this tool.
So a typical installation recipe might be to do:
## set up the environment
pip install virtualenv
virtualenv --no-site-packages --distribute curation
. ./curation/bin/activate
cd curation
## install the layercake branch of RDFLib and other prerequisites
pip install pyparsing
wget --no-check-certificate \
https://bitbucket.org/okfn/curate/raw/tip/pip-requirements.txt
pip install -r pip-requirements.txt
## this is needed for talking to SPARQL endpoints and cannot
## be installed with pip
easy_install 4Suite-XML
## install the curate tool
pip install curate