node-iceberg

A lightweight Node.js tree-based scraper/crawler. No more callbacks! Just async/await.

Installation

Download and install via npm package manager (stable):

npm install node-iceberg --save

Or clone fresh code directly from git:

git clone https://github.com/manuasir/node-iceberg.git
cd node-iceberg
npm install

Usage

This package allows to get filtered DOM elements from URLs throw customized iterators, and it works mainly in two ways:

• Scraper mode: Get elements using customized selectors.

// Example: download all links from a Blogspot URL. Use inside an 'async' function
const Iceberg = require('node-iceberg')
const getThisDomains = ['mediafire','mega','adf.ly'] // domains we can get
const conf = {
	// Iterator: Element that gets you to the next URL to process in blogspot
	iteratorElement: {
    	element: 'a',
    	cssClass: 'blog-pager-older-link' // optional
	},
	// Desired data to get extracted from the DOM. Example: Download links
	selector: {
    	element: 'a',
    	attrib: 'href',
    	values: getThisDomains // optional
	}
}
// Max Level Depth to explore: max blog pages
const maxLevelDepth = 10
const scraper= new Iceberg("http://someblog.blogspot.com")
const results = await scraper.start(maxLevelDepth,conf)
// or load a filter:
const scraper= new Iceberg("http://someblog.blogspot.com")
const conf = scraper.service('blogspot')
const results = await scraper.start(maxLevelDepth,conf)

Some websites are allowed to paginate directly from the URL using parameters. Ej: http://url?page=1,2,3... In that case, use it like this: pass only the configuration object and set max page within it:

// Example: get insecure cameras: insecam.org
const Iceberg = require('node-iceberg')
const conf = {
    iteratorElement: {
        url: url,
        iterator: '?page=',
        maxPage: 5 // maximum page to explore
    }, selector: { // elements we want to get
        element: 'img',
        attrib: 'src'
        }
    }
const scraper = new Iceberg("http://insecam.org")
const results = scraper.start(conf).then((results) => { console.log(results) }).catch((err)=>{ throw err })

• Crawler mode: Explores ALL links from a URL until the depth threshold is reached. Generates a tree from crawled data. Already explored ways are included only once.

// Warning! this can become a very costful task
const Iceberg = require('node-iceberg')
const crawler = new Iceberg('http://reddit.com')
const conf = crawler.service('crawler')
const maxLevelDepth = 2
const results = crawler.start(maxLevelDepth,conf)then((results) => { console.log(results) }).catch((err)=>{ throw err })

Test

If you downloaded this package from NPM, it's already tested. Otherwise you can test it like this:

npm test