Refactor and generalize functions

[Zivic]

• [ ] Handle missing data (memory) by accessing a public GPU api
• [ ] Ensure same type is used for both scraper's responses
• [x] Generalize backend "webscraper" function so it can be used for writing any data of this type to the DB
• [x] This also means generalizing the image upload function so it appends the correct source name

Originally posted by @Zivic in https://github.com/Zivic/GPU-price-tracker/issues/5#issuecomment-1465321536

[Zivic]

There is no public api, and the main website for gpu information has scraper protection This information includes: GPU Chip | Released | Bus | Memory | GPU clock | Memory clock | Shaders / TMUs / ROPs

• [ ] As we only need to get the data once, we can try manually copying the data
• [ ] This will be a separate "model" table with 1-n reference to the existing "product" table, Only difference should theoretically be the manufacturer
• [ ] Memory may need to be parsed
• [ ] Memory may need to be removed from "products"
• [ ] ! Find a way to compare the scraped name and match it to the correct "GPU Chip" entry
• [ ] The entire concept of the "Products" table may need to be redesigned. Manufacturer + Model = unique identifier, each of those can then have a specific price for each store