TwKit
正在载入
TwKit

I have a graveyard of dead scrapers. Every one died the same way: the site mov… · @ReignSwords · TwKit

@ReignSwords

I have a graveyard of dead scrapers. Every one died the same way: the site moved a div, my CSS selector went blind, and the pipeline handed me an empty CSV at 3am. Scrapling is the first framework I've seen that treats that as the actual problem, not a user error. It's a Python scraping library — fetchers, stealth browsers, spiders, the whole pipeline. Its standout is adaptive element tracking: instead of pinning ".price", it tracks the element itself. Site redesigns? It follows along. It also matches similar elements across a page, which is how most of my scrapers actually die — the tenth product card renders slightly different from the first nine. Three things I verified tonight: 1. The whole pipeline in one library — fast HTTP fetch, a stealth browser that handles Cloudflare Turnstile, parsing, concurrent spiders, JSON/CSV export. 2. It ships an MCP server and an agent skill. Your coding agent can browse and extract by itself. That's where this is all going. 3. Install is `pip install "scrapling[fetchers]"` — and yes, you need the [fetchers] extra. Plain `pip install scrapling` gives you a Fetcher that imports nothing. Learned that one the fun way. Five lines later I had 10 quotes off a live page in 1.6 seconds. That's the whole pitch: less time writing selectors, more time using data. Cold water: anti-bot bypass is an arms race, not a finished product — treat the Turnstile claims as "works today". And adaptive tracking is smart, not magic. If the entire page got rewritten in React, you'll still be debugging. But 96K stars and 1.6K new ones a day don't lie about how real this pain is. Repo's in the first comment. BSD-3 license, so your company lawyer won't flinch. If you scrape the web for a living, follow along — I dig up stuff like this every day. When was the last time your scraper survived a site redesign without a 3am fix? Be honest. 🔵 Fellow Premium members, let's connect! I post about AI & tech — follow for more 👥

曝光 136 · 评论 8 · 点赞 1 · 书签 0

查看原帖