Arachnid2 is a simple, fast web-crawler written in Ruby. You can use typhoeus to make HTTP requests, or Watir to render pages. bloomfilter-rb stores the URLs, and nokogiri finds the URLs on each webpage.

Arachnid2 is a successor to Arachnid, and was abstracted out of the Tellurion Bot.

The default use case for Arachnid2 is surfacing responses from a domains' URLs by visiting a URL, collecting any links to the same domain on that page, and visiting those to do the same.

The simplest way to use the gem is collecting all of the responses while spidering from some URL.

require "arachnid2"

url = "http://www.maximumfun.org"
spider = Arachnid2.new(url)
responses = []

spider.crawl { |response|
  responses << response
}

Obviously this could become unwieldy, so you can execute logic within the spidering to collect a narrow subset of the responses, transform or dissect the response, or both (or whatever you want).

require "arachnid2"
require "nokogiri"

url = "https://daringfireball.net"
spider = Arachnid2.new(url)
responses = []

spider.crawl { |response|
  responses << Nokogiri::HTML(response.body) if response.effective_url =~ /.*amazon.*/
  print '*'
}

Arachnid2#crawl will return always nil.

require "arachnid2"

url = "http://sixcolours.com"
spider = Arachnid2.new(url)
opts = {
  followlocation: true,
  timeout: 300,
  time_box: 60,
  max_urls: 50,
  :headers => {
    'Accept-Language' => "en-UK",
    'User-Agent' => "Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:47.0) Gecko/20100101 Firefox/47.0",
  },
  memory_limit: 89.99,
  proxy: {
    ip: "1.2.3.4",
    port: "1234",
    username: "sam",
    password: "coolcoolcool",
  }
  :non_html_extensions => {
    3 => [".abc", ".xyz"],
    4 => [".abcd"],
    6 => [".abcdef"],
    11 => [".abcdefghijk"]
  }
}
responses = []

spider.crawl(opts) { |response|
  responses << response
}

Tell Typhoeus to follow redirections.

Tell Typheous or Watir how long to wait for page load.

The crawler will time-bound your spidering. If no valid integer is provided, it will crawl for 15 seconds before exiting. 10000 seconds is the current maximum, and any value above it will be reduced to 10000.

The crawler will crawl a limited number of URLs before stopping. If no valid integer is provided, it will crawl for 50 URLs before exiting. 10000 seconds is the current maximum, and any value above it will be reduced to 10000.

This is a hash that represents any HTTP header key/value pairs you desire, and is passed directly to Typheous. Before it is sent, a default language and user agent are created:

The HTTP header Accept-Language default is en-IE, en-UK;q=0.9, en-NL;q=0.8, en-MT;q=0.7, en-LU;q=0.6, en;q=0.5, \*;0.4

The HTTP header User-Agent default is Mozilla/5.0 (Macintosh; Intel Mac OS X 10_13_4) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/11.1 Safari/605.1.15

Provide your IP, port for a proxy. If required, provide credentials for authenticating to that proxy. Proxy options and handling are done by Typhoeus.

This is the list of TLDs to ignore when collecting URLs from the page. The extensions are formatted as a hash of key/value pairs, where the value is an array of TLDs, and the keys represent the length of those TLDs.

In case you are operating the crawler within a container, Arachnid2 can attempt to prevent the container from running out of memory. By default, it will end the crawl when the container uses >= 80% of its available memory. You can override this with the option.

The crawler attempts to stop itself from returning data from links that are not indicative of HTML, as detailed in Arachnid2::NON_HTML_EXTENSIONS.

If you have setup a cache to deduplicate crawls, set a cached service url export ARACHNID_CACHED_SERVICE_ADDRESS=http://localhost:9000

This expects a push and get JSON API to respond to /typhoeus_responses, with a URL and the options pushed exactly as received as parameters. It will push any crawls to the service, and re-use any crawled pages if they are found to match.

Arachnid2 can crawl links with Watir, gathering up links like crawling with Typhoeus, but with pages that are actually rendered. You can access this option in one of two ways:

# ...
Arachnid2.new(url).crawl_watir(opts)
# -or-
with_watir = true  # the default is `false`
Arachnid2.new(url).crawl(opts, with_watir)

Arachnid2 has base defaults which you might want to address when employing Watir.

• First, the default crawl time is 15 seconds. As browser page loads can take this long, you will probably want to set a higher crawl time.
• Also, simply storing the browser is not a great idea, since it will be inaccessible after it is closed. Instead, consider nabbing the HTML, cookies, or whatever content is required during the crawl.
• Finally, note that Firefox is the default browser.

require 'arachnid2'

with_watir = true
responses = []
url = "http://maximumfun.org"
max = 60
browser = :chrome
opts = {time_box: max, browser_type: browser}

spider = Arachnid2.new(url)
spider.crawl(opts, with_watir) do |response|
  response.body.wait_until(&:present?)
  responses << response.body.html if response.body.present?
end

See the Typhoeus options above — most apply to Watir as well, with some exceptions:

Watir proxy options are formatted differently:

proxy: {
  http: "troy.show:8080",
  ssl: "abed.show:8080"
},

Proxy options handling is done by Watir.

And it accepts an argument to make browse headlessly

opts = { headless: true }

It accepts an argument mapped to Webdriver::UserAgent::Driver's agent option

opts = { agent: :desktop }

And it accepts an argument mapped to Webdriver::UserAgent::Driver's orientation option

opts = { orientation: :landscape }

These options do not apply to Watir, and will be ignored.

Fork the repo and run the tests

bundle exec rspec spec/

Bug reports and pull requests are welcome on GitHub at https://github.com/samnissen/arachnid2. This project is intended to be a safe, welcoming space for collaboration, and contributors are expected to adhere to the Contributor Covenant code of conduct.

The gem is available as open source under the terms of the MIT License.

Everyone interacting in the Arachnid2 project’s codebases, issue trackers, chat rooms and mailing lists is expected to follow the code of conduct.