This application will display gift recommendations for any Twitter user and is an example of what can be built using the Hunch API. The code is almost identical to that powering the Hunch Gift-o-Matic. The application is built to run either as a standalone Django application or using Google App Engine.
Other libraries used include:
- jQuery JavaScript Framework
- Mustache JavaScript templating library
- Hunch JavaScript SDK
- djangoappengine
-
Download the Google App Engine Python SDK
-
Test your app loads the test page
- run
sh dev_server.sh(this defaults to running on port 80 which may require root permissions) - open a web browser and navigate to http://localhost
- ensure that you see the success page
- now navigate to http://localhost/gifts/
- ensure that you can enter a twitter name and get recommendations
- run
-
When ready, deploy the app to GAE
- in settings.py, create a SECRET_KEY
python manage.py deploy
Start the development server
sh dev_server.sh
Publish your app to GAE
python2.5 manage.py deploy
Launch a local Python console for interecting with the app and datastore
python2.5 manage.py shell
Launch a remote Python console for interacting with the app and datastore
python2.5 manage.py remote shell
-
GAE will penalize your app if requests take over 1000 ms to complete, so push as many calls to the frontend as possible by using the Hunch Javascript SDK
-
This application is packaged with djangoappengine for creating Django projects that run on GAE. It is worthwhile to read the overview of using the helper.
-
GAE uses Python 2.5, so running scripts like
manage.pywith python2.5 is recommended. The scriptsdev_server.shandfresh_dev_server.shuse python2.5 -
You can access the GAE admin console at http://localhost/_ah/admin/
-
For testing, it is useful to redirect your-app-name.appspot.com to your local machine. To accomplish this, edit your /etc/hosts file and point that address to your IP (or localhost)