Collections

Collections is a prototype project at the Center for Open Science. This project is experimental, scope, technologies, code and functionality may change. This app has two main parts. The service stores data about the collection, and the client lets users interact with their collections.

Prerequisites

You will need the following things properly installed on your computer.

• Git
• Python3
• Postgresql
• Node.js (with NPM)
• Bower
• Ember CLI
• PhantomJS
• Yarn

Installation

Get the code:

$ git clone [email protected]:cos-labs/collections-service.git
$ cd collections-service

Install Dependencies:

Setting up a virtual environment for Python 3 is recommended.

$ pyenv virtualenv 3.6.0 osf-collections
$ pip install -r requirements.txt

Set Up Postgres:

OSX

$ brew install postgres
$ createdb
$ brew services postgres start

Set Up Redis

OSF Collections uses Redis as a database for celery tasks, as well as a cache.

OSX

$ brew install redis
$ brew services start redis

Set Up Tika

Tika is an apache product that allows for scraping text from a very large variety of file formats. Collections uses it to provide full-text search on documents that are created in collections.

OSX

$ brew install tika

Note this may require installing xcode command line tools.

Now that tika is installed, create a launch daemon to make starting and stopping tika more simple, and load it in so launchctl knows about it.

$ cp <collections-service>/apache.tika.plist /Library/LaunchDaemons/apache.tika.plist
$ launchctl load /Library/LaunchDaemons/apache.tika.plist

Now tika can be started and stopped using

$ launchctl start apache.tika
$ launchctl stop apache.tika

Set Up Email

Sendgrid

Simply export the Sendgrid token in the environment so Django can pick it up

export SENDGRID_API_KEY='SG.the_key_given_to_you_by_sendgrid'

Running

$ cd {collections}/

Set up local.py settings

(SA refers to "Social App", SU to Superuser)

SA_CLIENT_ID = "55b229c83f6946fe8c16b86217781028"
SA_CLIENT_SECRET = "70z9QBGWmYlFoXUr6HYZoi3QPr4ksafSEbJvTyNr"
SA_APPLICATION_NAME = "Collections"
SA_SITE_ID = 1
SA_PROVIDER_NAME = "osf"

SU_USERNAME = "admin"
SU_PASSWORD = "password"
SU_EMAIL = ""

Run migrations, create SocialApp login rewuirements

$ ./manage.py migrate
$ ./reset

Run the service

$ python manage.py runserver

Visit the api at http://localhost:8000/api/ or admin panel at http://localhost:8000/admin/.

Run the client

Follow the set-up instructions in the README for https://github.com/cos-labs/collections.

Visit your app at http://localhost:4200.

Configuration

Set up the backend to use either staging or prod. Do this in the sessions for both the client and the service. export BACKEND=prod or export BACKEND=stage

Create a developer app at https://staging.osf.io/settings/applications/ with the following settings:

• Project homepage URL: http://localhost:8000/
• Callback URL: http://localhost:8000/accounts/osf/login/callback/

Create a django super user:

$ python manage.py createsuperuser

Create a new Site with domain name: http://localhost:8000/ in the django admin panel:

• Note: The site id must match the SITE_ID variable defined in SITE_ID should be set to the site id in collections-service/service/settings/local.py
• The SITE_ID can be found in the URL
• You may have to define SITE_ID in the local.py file

Create a SocialApplication in the django admin panel:

• Set provider to "Open Science Framework"
• Set the client id and secret key to the ones defined in your developer app
• Select http://localhost:8000/ as the site

Build the search index

• python manage.py rebuild_index
• This needs to be done on first run / population and then on a semi-regular basis when running the API to index new content from the API.

Populating the database

• Before moving forward, make sure you've logged into the test server using your OSF staging credentials.
• To populate the database with users, meetings, collections, and items, run $ python populate.py from the project root.

Running Tests

• python manage.py test