Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
Since I originally wrote this a module called request has come on the scene. You might want to try that before mucking about with extractor-js. A small NodeJS package using jsDom to facilite screen scraping and spidering. It scrapes single and multiple elements and includes support for many tag attributes.
JavaScript
Branch: master

Merge pull request #25 from touchlocal/205ac44123186ebf27e588ab9f36e9…

…62878e46f6

Allow extractor-js to follow redirects
latest commit 76468daaa5
@rsdoiel authored

README.md

build status

extractor-js

FORK NOTE!

This is a fork of https://github.com/rsdoiel/extractor-js, but adds the ability to follow 302 redirects by plugging in https://npmjs.org/package/follow-redirects module.

Before you dive in

When I originally wrote extractor-js there wasn't allot of libraries that did what I needed at the time. Since then a very solid library with lots of community usage called request has come along. You should consider it before choosing extractor. The URL is https://github.com/mikeal/request. Otherwise, enjoy extractor.

Overview

Periodically I wind up writing scripts to screen scrape or spider sites. Node is really nice for this. extractor-js is a utility module for these types of script.

Four methods

extractor has four methods -

  • fetchPage - get a page from the local file system or via http/https
  • scrape - A combination of FetchPage(), Cleaner(), and Transformer() which fetches the content, parses it via jsDOM/jQuery and extracting interesting parts based on the jQuery selectors pass to it creating an object with the same structure as the selector object and passing it to Scrape()'s callback method. You can override the Cleaner() and Transformer() methods by passing in your own cleaner and transformer functions.
  • spider - a utility implementation of Scrape using a fixed map for anchors, links, scripts and image tags.

Example (Scrape)

var extractor = require('extractor'),
pages  = [ 'http://nodejs.org', 'http://golang.org'],
selector = {
    'page_title':'title',
    'titles' :'h2',
    'js' : 'script[type=text/javascript]',
    'css' : 'link[type=text/css]',
    'about_as_id' : '#about',
    'about_as_class' : '.about'
    };

pages.forEach(function(page) {
    extractor.scrape(page, selector, function (err, data, env) {
        if (err) throw err;

        console.log("Processed " + env.pathname);
        console.log("Last modified " + env.modified);
        console.log("Page record: " + JSON.stringify(data));
    });
});

This example script would process three pages from the pages array and output a console log of the processed page and a JSON representation of the content described by the selectors.

Example (Spider)

In this example we spider the homepage of the NodeJS website and list of the links found on the page.

var extractor = require('extractor');

extractor.spider('http://nodejs.org', function(err, data, env) {
    var i;
    if (err) {
        console.error("ERROR: " + err);
    }
    console.log("from -> "+ env.pathname);
    console.log("data -> " + JSON.stringify(data));
    for(i = 0; i < data.links.length; i += 1) {
        console.log("Link " + i + ": " + data.links[i].href);
    }
});
Something went wrong with that request. Please try again.