Python/Django reimplementation of tatoeba.
See requirements.txt.
You can choose your graphing backend to either be pure python (networkx) or C (scipy). The python backend should work readily with any flavor of python. The C one might not but is insanely fast on CPython. To install the C backend make sure you have gcc, gfortran, python-dev, libblas-dev, liblaplack-dev, and python-pip. Also you probably will need > 1GB RAM for compiling the dependencies.
You can also choose and configure a search engine backend for haystack. Consult haystack's docs for more information. The template project provides a configuration for xapian. The dependencies is basically the header files libxapian-dev and the python bindings python-xapian. To install it inside a virtualenv install libxapian-dev on your system and then run the provided xapianvenv.sh script from within the virtualenv.
Then in a terminal:
pip install -r requirements.txt
For the graphing backend:
pip install -r scipy\_backend\_requirements.txt
or
pip install -r networkx\_backend\_requirements.txt
For the xapian search backend:
pip install -r xapian\_backend\_requirements.txt
See the pytoeba repository for a fully configured project using pytoeba.
In the root directory of the configured project run:
python manage.py syncdb --migrate
python manage.py runserver
It should be accessible now on 127.0.0.1:8000
To run the test suite:
py.test
To see the coverage report:
coverage run --source pytoeba -m py.test
coverage report
To generate an html report:
coverage run --source pytoeba -m py.test
coverage html
Then check htmlcov/index.html in your browser for a line by line coverage report per module