I’m referring to projects like redlib or invidious.
I was thinking about doing something similar for a local second-hand marketplace and got curious. Redlib seems to use token spoofing to get past rate limits and Invidious doesn’t even use the official YouTube API.
The only way I thought of, which would be slow, is to scrape the site (like you would with Beautiful Soup).
Well, it’s as slow as the website you’re scraping. Could actually be faster if you don’t have to execute a lot of bullshit JavaScript. And for the rest clever caching should help.
In terms of technology you’re looking for XSLT, Xpath, CSS selectors and whatever parsers are available for your language of choice. Don’t ever attempt to use regex for scraping.
seconded
Proceeds to implement a scraper in bash with grep and awk because it’s the only way i know how to
You’ve given me a great jumping off point, thank you!