Fork this project to your GitHub account.
Clone your version of the project into your local repository.
Then Copy the file
kikar_hamedina/kikar_hamedina/local_settings.py.template into
kikar_hamedina/kikar_hamedina/local_settings.py
Edit it to include your facebook details (Ask project leader for details)
- FACEBOOK_SECRET_KEY
- FACEBOOK_APP_ID
Install virtualbox: https://www.virtualbox.org/wiki/Downloads
Install vagrant: https://www.vagrantup.com/downloads.html
cd to kikar_hamedina root directory
Run vagrant up - this can take about 25-30 minutes
For a successful run you should see something like 'kikar start/running, process 10031' in the end.
This means a virtual machine was successfully created by vagrant (it's an Ubuntu server) and a Django server is running on it on port 8000. The virtual machine and HTTP server are connected to the files on your environment so you can use you favorite text editor to change any project files and these will be reflected on the server (Django will apply changes automatically and reload the server). You can now access the server from your browser on http://localhost:8000/
Virtual machine (vagrant) maintenance:
vagrant destroy - destroy virtual machine
vagrant up - create virtual machine or run it if it is down
vagrant reload --provision - re-run the startup script to get the machine and server up and running and apply changes that were made, update data from facebook etc.
vagrant global-status - see status of all machines
vagrant ssh <id> - ssh into machine
After SSH'ing into the machine, you can see the server log in /var/log/upstart/kikar.log (e.g. sudo tail -F /var/log/upstart/kikar.log)
These commands can be used to manage the django server (these are standard Upstart commands):
sudo stop kikar
sudo start kikar
sudo restart kikar
sudo status kika
.. Set the following environment variables:
- DJANGO_SETTINGS_MODULE (for local version set 'kikar_hamedina.settings.local')
- SECRET_KEY
- FACEBOOK_SECRET_KEY
- FACEBOOK_APP_ID
.. Set your db inside kikar_hamedina/kikar_hamedina/settings/local.py with your db details.
.. Install required packages: pip install -r requirements/local.txt (for local version requirements)
To Set up DB run the following:
python manage.py syncdb --migrate
python manage.py runserver
This is all very nice, but there's no data :-( To insert initial data into the db, CTRL-C the server abd run:
python manage.py loaddata data_fixture_planet.json
python manage.py loaddata data_fixture_mks.json
python manage.py loaddata data_fixture_persons.json - This is an empty json, waiting for deprecation
python manage.py loaddata data_fixture_facebook_feeds.json
python manage.py fetchfeedproperties
python manage.py fetchfeedstatuses --initial
The first command will insert to db data from a pre-created list of parties, persons, feeds, and tags. The second command will fetch statuses for all feeds in data. Tag initial sets the request to 1000 messages per feed (as opposed to default of 20), so it might take a while to finish. *Note: Default number of messages is set within fetchfeedstatuses.py command, see file for exact value.
You can also run: python manage.py fetchfeedstatuses {feed_id} [--initial] to download a single feed.
Once you're all setup run:
python manage.py runserver
and enjoy the show.
At the main directory there's a sub-directory called data. Within it there are four csv files, for Party, Person, Feed, and Tag data, which can be edited.
- Note that in order to handle encoding correctly those csv files are UTF-8, which Microsoft Excel does not create as default. You may reffer to http://stackoverflow.com/questions/4221176/excel-to-csv-with-utf8-encoding for help.
Once you're done editing, create the json file by running data_fixture_helper_script_csv_to_json.py, and replace the old data_fixture_facebook_feeds.json with the new one.
Another option would be to edit data with the django Admin, and then run:
python manage.py dumpdata core.Party core.Person core.Facebook_Feed core.Tag --indent 4 > <file_name>.json
A script to convert the json to the csv files exists as well. Note that currently the file names are hardcoded, so use with caution..