There are many shopify scrapers out there but I doubt any of them are using multithreading. With this in mind, I thought I could use what I learned in ECE 252 at the University of Waterloo to build a shopify scraper that outperforms existing ones.
I have modelled this program as a modified producer-consumer problem. I say modified because the shared buffer does not have an upper bound on the number of spaces available. Instead, only one general semaphore, items, is required, which keeps track of the number of full spaces in the buffer. When a producer thread has done it's job and pushed the result to the buffer, it will post on items. At the same time, consumer threads will be waiting on items so that they can consume what was produced.
Each producer thread handles a different page of the /products.json?page=X endpoint.
Run make to create the executable main file.
Run ./main <BASE_URL> where <BASE_URL> should be replaced with a shopify website base url. For example,
./main http://example.com.