OMSCS_Course_Availability

Scripts to make navigating OSCAR registration easier

File overview

• coursexp.py - Course Explorer - library used to interact with OSCAR
• etracker.py - Enrollment Tracker - Stores enrollment info in a database for later reference
• regpage.py - Navigate directly to registration page. It doesn't get much faster than this.

Environment variable setup

To use the scripts, at minimum your GT username and password will need to be supplied. To prevent putting your id/password in your history and to avoid accidentally commiting them to git these scripts expect the use of an environment variables file. Necessary DUO login actions still need to be handled separately.

To setup the .env file:

1. In the same directory as your OMSCS_course_availability.py, open a .env with your favorite editor

vim .env

2. Add your login info to the file

OMS_ID=yourUserName
OMS_PWD=yourPassword

3. Save

Additionally, etracker currently expects the following in the .env file:

• EMAIL_PWD - password for smtp server login
• EMAIL_USER - username for smtp server login
• TO_EMAIL - email address messages should go to
• FROM_EMAIL - email address messages should be reported as from

Requirements

Python requirements in requirements.txt
System requirements outlined in Dockerfile

However, the basic needs are:

0. Python 3.x
1. Browser (Firefox)
2. geckodriver for Firefox
3. Python requirements (install with "pip install -r requirements.txt")
   1. lxml
   2. selenium
   3. apscheduler
   4. python-dotenv

Using other browsers should work as well, but the appropriate driver will be needed.
To "install" the driver, add it to your path. On Linux you can place the file in "/usr/local/bin"
Additionally, the line identifying the browser will also need to be edited: "browser = webdriver.Firefox()"

Docker

Build

In the directory containing the Dockerfile, run:
docker build -t <newimage_name> .

Run

docker run -d -v /local/directory/for/data:/shared/container/directory <image_name> python3.6 etracker.py

-d - detached mode; runs as a background process
-v - map volume

Both flags are very important here.
If -d isn't used, the container will close immediately and nothing will be run.
If -v isn't used, if the process ever stops running, all generated data will be lost.

Share image

(Not needed if run from same system build is preformed on)

0. Save file
   docker save -o <output_img.tar> <image_name>
1. Transfer file (eg w/ scp)
   scp <output_img.tar> user@address:/dest/directory
2. Load image into docker in new env
   docker load -i <output_img.tar>
3. Verify image present
   docker images

TODOs

• Add argparser to etracker
• Add check for new semester being made available. (Need to catch first students added. Missed for Spring 2019)
• Migrate to Python 3.7.1+
• dbadd.semester_prefix -- needs to be dynamic for each semester
• Add support for multiple course types, especially CSE. Currently pinned to only online CS courses. Add CLI flag.
• Get root cause for daily "unspecified errors" being logged.