Web Crawler is a bot that downloads the content from the internet and indexes it. The main purpose of this bot is to learn about the different web pages on the internet. This kind of bots is mostly operated by search engines. By applying the search algorithms to the data collected by the web crawlers, search engines can provide the relevant links as a response for the request requested by the user. In this article, let’s discuss how the web crawler is implemented.
Webcrawler is a very important application of the Breadth-First Search Algorithm. The idea is that the whole internet can be represented by a directed graph:
- with vertices -> Domains/ URLs/ Websites.
- edges -> Connections.
Example:
Approach: The idea behind the working of this algorithm is to parse the raw HTML of the website and look for other URL in the obtained data. If there is a URL, then add it to the queue and visit them in breadth-first search manner.
Note: This code will not work on an online IDE due to proxy issues. Try to run on your local computer.
Java
// Java program to illustrate the WebCrawler import java.io.BufferedReader; import java.io.InputStreamReader; import java.net.URL; import java.util.HashSet; import java.util.LinkedList; import java.util.List; import java.util.Queue; import java.util.regex.Matcher; import java.util.regex.Pattern; // Class Contains the functions // required for WebCrowler class WebCrowler { // To store the URLs in the / /FIFO order required for BFS private Queue<String> queue; // To store visited URls private HashSet<String> discovered_websites; // Constructor for initializing the // required variables public WebCrowler() { this .queue = new LinkedList<>(); this .discovered_websites = new HashSet<>(); } // Function to start the BFS and // discover all URLs public void discover(String root) { // Storing the root URL to // initiate BFS. this .queue.add(root); this .discovered_websites.add(root); // It will loop until queue is empty while (!queue.isEmpty()) { // To store the URL present in // the front of the queue String v = queue.remove(); // To store the raw HTML of // the website String raw = readUrl(v); // Regular expression for a URL String regex // To store the pattern of the // URL formed by regex Pattern pattern = Pattern.compile(regex); // To extract all the URL that // matches the pattern in raw Matcher matcher = pattern.matcher(raw); // It will loop until all the URLs // in the current website get stored // in the queue while (matcher.find()) { // To store the next URL in raw String actual = matcher.group(); // It will check whether this URL is // visited or not if (!discovered_websites .contains(actual)) { // If not visited it will add // this URL in queue, print it // and mark it as visited discovered_websites .add(actual); System.out.println( "Website found: " + actual); queue.add(actual); } } } } // Function to return the raw HTML // of the current website public String readUrl(String v) { // Initializing empty string String raw = "" ; // Use try-catch block to handle // any exceptions given by this code try { // Convert the string in URL URL url = new URL(v); // Read the HTML from website BufferedReader be = new BufferedReader( new InputStreamReader( url.openStream())); // To store the input // from the website String input = "" ; // Read the HTML line by line // and append it to raw while ((input = br.readLine()) != null ) { raw += input; } // Close BufferedReader br.close(); } catch (Exception ex) { ex.printStackTrace(); } return raw; } } // Driver code public class Main { // Driver Code public static void main(String[] args) { // Creating Object of WebCrawler WebCrowler web_crowler = new WebCrowler(); // Given URL String root // Method call web_crowler.discover(root); } } |
Output:
Website found: https://www.google.com Website found: https://www.facebook.com Website found: https://www.amazon.com Website found: https://www.microsoft.com Website found: https://www.apple.com
Problem caused by web crawler: Web crawlers could accidentally flood websites with requests to avoid this inefficiency web crawlers use politeness policies. To implement politeness policy web crawler takes help of two parameters:
- Freshness: As the content on webpages is constantly updated and modified web crawler needs to keep revisiting pages. For that freshness uses HTTP protocol to as HTTP has a special request type called HEAD which returns the information about the last updated date of webpage by which crawler can decide the freshness of a webpage.
- Age: An age of a webpage is T days after it has been last crawled. On average webpage updating follow Poisson distribution and the older a page gets the more costs to crawl the web page so Age is more important factor for crawler than freshness.
Applications: This kind of web crawler is used to acquire the important parameters of the web like:
- What are the frequently visited websites?
- What are the websites that are important in the network as a whole?
- Useful Information on social networks: Facebook, Twitter… etc.
- Who is the most popular person in a group of people?
- Who is the most important software engineer in a company?