Hacker News Reader

Comment by jasode | Hacker News Reader

It would be helpful if the documentation compared how Scraperjs is different from, or better than, CasperJS for scraping. CasperJS is the older and more well-known wrapper around PhantomJS so comparisons would help people decide what the appropriate tool would be.

http://casperjs.org/

View on HN

I guess the biggest difference is, that Casper isn't a NodeJS module. Interaction with node (using the archive of npm) becomes hard. I am releasing something similar in a couple of weeks, but aiming at js sandboxing instead of scraping. :)

For instances where you don't need a full featured browser to get the data you need, Scraperjs using the Cheerio backend should be WAY faster than casper/phantom.

I've not used Scraperjs yet, but cheerio is pretty great.

I think you'd need CasperJS when you need to perform browser actions (login, click on a particular button, fill a form etc). But if you just want to scrape content (eg episode urls of The Daily Show from Hulu), then ScraperJS should be enough (and faster?)

As far as I can see it has not much in common with casperjs, apart from the fact that it can use phantomjs.

If you mean that the syntax is different, yes, I get that.

CasperJS can also scrape dynamic websites. What criteria would someone want to use ScraperJS instead of CasperJS for that task? Are there features in ScraperJS that don't exist in CasperJS? Does it take 10x less lines-of-code to accomplish the same task? Etc.

For web scraping purposes ScraperJS and ScraperJS would probably use the same lines of code, however ScraperJS has most of the tools you need for web scraping, something that CasperJS lacks (that's not their main goal). ScraperJS is move flexible than just CasperJS. If you want static content just use the static scraper and get lightening fast results. TL;DR: CasperJS is great but it's not made for web scraping.

Reply on news.ycombinator.com