Reporta is your personal reporter. Bringing a conversational interface to news. AlJazeera Media in Context Hackathon.

Make sure you have python-dev installed in order to build numpy. On ubuntu it is:

apt-get install python-dev

You need to activate the local environment to use the python virtualenv

source server/venv/bin/activate

Then if it is your first install you should do a pip install to get the requirements

pip install -r server/requirements.txt

You also need redis installed and running. Install docker first, and then install redis on docker with.

.server/redis/install.sh

You also need to download the NLTK files and build the NLP class.

python server/nltk/install.py
python server/nltk/build.py

First run Redis with

./server/redis/start_redis.sh

Then run the server with

python server/server.py

Sometimes you want the data in Redis to persist between runs, such as if you are running Reporta in production.

First you need a data Dockerfile, such as the one here:

[https://github.com/lingz/data-docker](Data Docker)

If you don't have this, then download and run it as so

./server/redis/install_data.sh
./server/redis/start_data.sh

Once the redis-reporta data docker container exists, you can start redis with the persist command line option

./server/redis/start_redis.sh persist

Then start the server as normal

python server/server.py

This project is licensed under the MIT license, found in LICENSE.md