The server will be written in Python. While Rodrigo Moraes’ work on tipfy and webapp-improved look interesting, we anticipate using webapp. The server will:
- use a scheduled task to initiate a URL Fetch
- pull data from a remote source with URL Fetch, save it to a datastore, and enqueue a task onto the task queue
- push tasks from the task queue to an application URL which will process the data and store the result in the datastore
- offer a RESTful interface to the result of the processed data through Google ProtoRPC
- rely on the Users API and app.yaml for authentication and authorization
The client will be written in JavaScript and use the Angular framework. It will:
- have resources that pull JSON from the Google ProtoRPC services
- display the data with @ng:repeat
- use Google Prediction API and Google Chart Tools to analyze and vizualize the data
- use Closure for controllers if appropriate