course-scheduler

Website

This is a website I made to be a resource for the students of Stevens Institute of Technology

Future plan: Use topological search to make a study plan creator! Given a list of courses, possibly gotten from parsing one of the pages at the school website, it would supply one or possibly more study plans, taking into account prereqs and coreqs.

Background

So at Stevens the only way to reasonably figure out your schedule for future semesters is to use the course scheduler available here. You can search for classes and see different offerings, and work out a schedule that suits you best.To figure out your schedule you basically just need to use trial and error until you create one that works and you're relatively happy with. I want to make it so that you can enter the courses you need to take, and all possible schedules will be generated.

That's what inspired me to make a tool to see all possible schedules you could have, given the classes you want to take. Then I wanted to make it a web app, then it inherited the goal of another website I had to host course info.

Approach

So using the scheduler API, you can request the XML for any semester. I take the XML and a list of courses and pull out the info I need, and turn it into a big nested dictionary. Then I go through all the courses and sections, find all possible combinations, and check for conflicts. Any schedules with no conflicts are put into a new dictionary, and sent through flask to make the HTML to display for the user.

Hosting/tech

I'm using pythonanywhere to host the site and handle the server stuff. That's where anything from this repo goes. Then I'm using MacHighway to host the static, non-HTML files, mainly the stylesheet and any pdfs.

Python libraries

flask - the web framework that kinda runs everything

json - Course combos are stored as JSON in a cookie

os - to remove files that are temporarily saved (like the xml from the Scheduler API)

random - to get a random number as a temporary password for the admin area

smtplib - to text me the password for the admin area

hashlib - to encrypt the password for the admin area

pickle - to store course info as a list of class objects

urllib - to get the xml from the scheduler api

xml.etree.ElementTree - to turn the xml into something python can work with

re - regex matching used in xml parsing

itertools - to find all possible course combos

Key files

run.py - the main file for flask - runs the website

course_class.py - a python class to represent a course

scheduler.py - the script to get feasible schedules

secrets.py - the secret stuff to keep stuff secure

settings.py - some macro-type-stuff to control how the app works

templates/ - all the HTML files

Important processes

Parsing XML - The XML file is iterated over in different ways to pull out the info I need to schedule the courses, and all the info is saved as a (pretty nested) dictionary

Scheduling - Given the list of courses to be scheduled, all possible combinations are found. Then each combination is checked against the dictionary from "Parsing XML" to find conflicts. If no conflicts are found, the combination is added to a list of good combos

Course object - all courses for the Courses page are stored in the class defined by the course_class module. Department, number, name, lecture, recitation, lab, homework, exams, and final are attributes, books are a dictionary with the key/value pair file_name/name, where file_name is the name of the file in the location specified in the getHTML() method of the class.

Data storage - The xml file is not stored, it is saved temporarily, parsed and deleted. This allows each schedule to always have the most recent info. The course info is stored in courses.dat. All the Course objects are stored in a list, then that list is saved to (and loaded from) the file through pickle. The functions save_data(data) and load_data() are in the course_class module.

Contributing

Clone the repo, make a secrets.py with the following:

def send_message():
    login={}
    login['emailUsername'] = "USERNAME@gmail.com"
    login['emailPassword'] = "PASSWORD"
    login['phone_number']  = '##########@vtext.com'
    return login
def app_secret():
    return "Some_secret_thing_for_the_app"

**Optional(but suggested):**Make a virtualenv for the project. virtualenv venv, source venv/bin/activate, then pip install -r requirements.txt. this should install everything you need with the right versions. A virtualenc also lets you use different versions than you have installed globally. pip freeze will printout the current pip installs

You can exit the virtualenv with deactivate

Then you should be able to run it locally by going to the main directory and running python run.py and then pointing your browser to 127.0.0.1:5000. The stylesheet is pulled from my server, but you can easily add your own or modify mine by making a new file and changing templates/base.html to use the new one.

Mess around on your own, or add or fix something and create a pull request.