queens-hacks / qcumber-frontend

The web front-end for qcumber.ca

Qcumber Frontend

Note: If you're looking for the easy to use course catalog for Queen's University, you came close! This is the code for the frontend interface that powers it. You'll want to head to https://qcumber.ca for the end user site :).

This is a simple flask application which talks to Elasticsearch as a data store, and consumes the data found in qcumber-data.

Setup

Currently, the frontend is in a bit of a messy state, as it was thrown together over the course of a week to get it ready in time for course selection.

Clone this repository

$ git clone https://github.com/Queens-Hacks/qcumber-frontend

Create a virtualenv and install the requirements from the requirements.txt file into it.

$ virtualenv env
$ . env/bin/activate
$ pip install -r requirements.txt

Install elasticsearch and get it running. Currently the code only supports the default ports, but there will be options to configure that soon.

Clone the data repository into the out directory.

$ git clone https://github.com/Queens-Hacks/qcumber-data out

Run the fill.py script to load the data into the elasticsearch instance.

$ ./fill.py

Now the debug server can be run by running main.py

$ ./main.py

qcumber-frontend's Issues

Add uservoice integration for feedback

The original Qcumber site has uservoice integration for feedback. Would probably be useful to carry it over.

https://qcumber.uservoice.com/forums/181738-general

Minify JS/CSS

Currently we don't minify JS and CSS before they're sent to the client. This should probably happen.

Classfind links

cross post from ChrisCooper/QcumberD#98

They requested that we have links to classfind.com to help people find classes. Marking this as an enhancement because it would be a good additional feature.

Exam data needed

This is a reference to Queens-Hacks/qcumber-scraper#16 so that it can be included in the 1.0 milestone

Courses names should be links in the requirements text

This could be data that is pre-parsed, or just use a regex like the original site. The first method would be better.

Cache-bust Assets

Currently, if an asset updates, it is cached for basically forever. This is good for normal operation, but bad for upgrades. We should cache-bust the assets with a hash such that when the code updates, clients get the new code.