Articles can be downloaded and resumed

[todorus]

PR for #16

• Added the SqlAlchemy model that is used in the hunter-gatherer repo, so we both use the same model (I need to move this to a seperate repo, so it can be a seperate pip library instead)
• Articles can be streamed to a csv file by using data.collect.download_articles_to_file
• Downloads can be resumed when they are interrupted
• The word2vec model generation can load the data from a csv file in a streaming manner

example of how to download articles

from data import model
from data import collect

engine = model.engine(
    database_user=database_user,
    database_password=database_password,
    database_host=database_host,
    database_port=database_port,
    database_name=database_name
)

session = model.session(engine)

collect.download_articles_to_file(session, "articles.csv")

Then afterwards you can run python createModel.py input.csv output.w2vec to generate the model

[Tilana]

Great! Some things: MAybe we can find a better name for the data module. Data is too generic and often used as a folder to save raw data files...

[todorus]

Instead of a 'data' folder, what about 'datastore'?

[Tilana]

datastore is perfect!