alirezamika/autoscraper — reverse-engineered prompt

Build me a simple Python package that makes web scraping feel almost automatic. I want to give it either a page URL or raw HTML, plus a few example values I care about, like a title, a price, a link, or some text from the page. It should figure out the scraping pattern from those examples and then return matching items from that page.

After it learns once, I want to reuse that scraper on similar pages. It should support two main modes, one that finds similar results on a page, like related post titles, and one that gets the exact fields in the same order as my sample list, like stock price and market cap on another ticker page. Please let me pass normal request options too, like headers or proxies.

Also add a way to save the learned scraper to disk and load it later. Keep it lightweight and easy to use from a few lines of Python, with a couple of clear examples and basic tests. If you need to, look up current docs online.