Test scraping across years

[lhvy]

We need to make sure that Notangles works across the year boundary (mainly for T3 -> summer term + T1). This involves guaranteeing that our scraper (and database) can handle multiple years of data.

Tasks involved:

• Look at current GraphQL API structure and agree on a good way to represent multiple years of data
• Note any changes that will be made and ensure Notangles frontend and backend are updated accordingly
• Ensure scraper can handle scraping/storing multiple years and that it formats the data appropriately as per previous point
• Ensure data is added to API as per previous steps

[fritzrehde]

Currently, the course id is {course-code}{career}.
I think we should change it to {course-code}-{career}-{year}.
The reason we aren't including the {term} in the course id is that the handbook only contains one page per course per year, which contains data for the course that year. We use the same format to reduce duplicated data.

Currently, the class id is {course-id}-{class-nr}-{term}-{year}.
I think we shouldn't just blindly reuse the course id (advantages: 1. we can pick the exact fields that are needed for uniqueness, though that wouldn't really change anything here, 2. local reasoning in code is improved, 3. we can customize delimiters separately), so change class to {course-code}-{career}-{year}-{term}-{class-nr}.

Currently, the time id is {class-id}{day}{location}{time}{weeks}, where fields like time and weeks already contain dashes, so I propose using underscores as the field delimiter.
I think we should change the timing id to {course-code}_{career}_{year}_{term}_{class-nr}_{location}_{day}_{time}_{weeks}.