Itemize

A lazy, fluent web crawler with an async/await API.

$ yarn add itemize

Quickstart

Itemize lists all of the linked files and pages underneath the specified root URL.

const urls = itemize('https://news.ycombinator.com', { depth: 2 })

// Get a quick Hacker News sitemap
while (!urls.done()) {
  console.log(await urls.next())
}

This is useful for writing mirrors, monitoring a page for new content, etc. It starts at the root URL provided and automatically spiders through to find connecting pages. Itemize takes a lazy approach to I/O and only makes requests when you're asking it for more content with next().

API

itemize(url, options)

Returns an Itemize instance.

• url: String, the root URL from which to crawl
• options: Object
  • depth: Number, crawl this many layers deep (0)

const items = itemize('https://nodejs.org/download/release/', { depth: 1 })

.next()

Returns a Promise for a String, the next linked URL.

If no urls remain, returns a Promise for undefined.

const url = await items.next()

.done()

Returns a Boolean representing whether or not all spidering routes have been exhausted.

if (items.done()) console.log('crawl complete')

.all()

Returns a Promise for an Array of Strings, all of the previously traversed items.

const all = await items.all()

.close()

Itemize uses a keepalive HTTP/HTTPS agent. Use close() to destroy the existing underlying socket and create a new Agent with no existing connections.

You should use this to clean up after Itemize instances that haven't completed their crawls.

items.close()

Tests and Examples

$ yarn test

$ node --harmony examples/hackernews.js
$ node --harmony examples/nodes.js