skim

Scrape websites simply. Streaming HTML parser combined with a flexible HTTP client.

npm install skim
6 downloads in the last week
10 downloads in the last month

skim

Parse a stream of HTML using CSS selectors to extract data. Doesn't require a DOM, thus it's fast and low-memory.

Example

Define your scraping parameters in a JSON manifest:

rem.stream("http://news.ycombinator.com/").pipe(skim({
  "$query": "td.title ~ td ~ td.title > a",
  "$each": {
    "title": "(text)",
    "link": "(attr href)"
  }
}, function (err, json) {
  console.log(json);
});

Result:

[ { link: 'https://www.hackerschool.com/blog/5-learning-c-with-gdb',
    title: 'Learning C with gdb' },
  { link: 'http://blogs.scientificamerican.com/guest-blog/2012/08/27/the-hidden-truths-about-calories/',
    title: 'Hidden Truths about Calories' },
  { link: 'http://cantada.ca/',
    title: 'Can\'tada - Tracking the stuff you can\'t use in Canada' },
  { link: 'https://blog.gregbrockman.com/2012/08/system-design-stripe-capture-the-flag/',
    title: 'Seccuring Stripe\'s Capture the Flag' },
  { link: 'http://swanson.github.com/blog/2012/08/27/move-your-feet.html',
    title: 'Move your feet' },
    ... ]
npm loves you