Introduction to Web scraping with Java

Web scraping or crawling is the fact of fetching data from a third party website by downloading and parsing the HTML code to extract the data you want.

Since every website does not offer a clean API, or an API at all, web scraping can be the only solution when it comes to extracting website information.
Lots of companies use it to obtain knowledge concerning competitor prices, news aggregation, mass email collect…

Almost everything can be extracted from HTML, the only information that are “difficult” to extract are inside images or other medias.

In this post we are going to see basic techniques in order to fetch and parse data in Java.

Prerequisites

  • Basic Java understanding
  • Basic XPath

Tools

You will need Java 8 with HtmlUnit

<dependency>  
    <groupId>net.sourceforge.htmlunit</groupId>
    <artifactId>htmlunit</artifactId>
    <version>2.19</version>
</dependency>

If you are using Eclipse, I suggest you configure the max length in the detail pane (when you click in the variables tab ) so that you will see the entire HTML of your current page.

Let's scrape CraigList

For our first exemple, we are going to fetch items from CraigList since they don't seem to offer an API, to collect names, prices and images, and export it to JSON.

First let's take a look at what happen when you search an item on CraigList. Open Chrome Dev tools and click on the Network tab :

The search URL is :

https://newyork.craigslist.org/search/moa?is_paid=all&search_distance_type=mi&query=iphone+6s  

You can also use

https://newyork.craigslist.org/search/sss?sort=rel&query=iphone+6s  

Now you can open your favorite IDE it is time to code. HtmlUnit needs a WebClient to make a request. There are many options (Proxy settings, browser, redirect enabled ...)

We are going to disable Javascript since it's not required for our example, and disabling Javascript makes the page load faster :

String searchQuery = "Iphone 6s" ;

WebClient client = new WebClient();  
client.getOptions().setCssEnabled(false);  
client.getOptions().setJavaScriptEnabled(false);  
try {  
  String searchUrl = "https://newyork.craigslist.org/search/sss?sort=rel&query=" + URLEncoder.encode(searchQuery, "UTF-8");
  HtmlPage page = client.getPage(searchUrl);
}catch(Exception e){
  e.printStackTrace();
}
}

The HtmlPage object will contain the HTML code, you can access it with asXml() method.

Now we are going to fetch titles, images and prices. We need to inspect the DOM structure for an item :

With HtmlUnit you have several options to select an html tag :

  • getHtmlElementById(String id)
  • getFirstByXPath(String Xpath) -getByXPath(String XPath) which returns a List
  • many others, rtfm !

Since there isn't any ID we could use, we have to make an Xpath expression to select the tags we want.

XPath is a query language to select XML nodes( HTML in our case).

First we are going to select all the <p> tags that have a class `result-info

Then we will iterate through this list, and for each item select the name, price and url, and then print it.

List<HtmlElement> items = (List<HtmlElement>) page.getByXPath("//p[@class='result-info']" ;  
if(items.isEmpty()){  
  System.out.println("No items found !");
}else{
for(HtmlElement item : items){  
  HtmlAnchor itemAnchor =  ((HtmlAnchor) htmlItem.getFirstByXPath(".//a"));

  String itemName = itemAnchor.asText();
  String itemUrl = itemAnchor.getHrefAttribute() ;
  HtmlElement spanPrice =((HtmlElement) htmlItem.getFirstByXPath(".//span[@class='result-price']")) ;
  // It is possible that an item doesn't have any price
  String itemPrice = spanPrice == null ? "no price" : spanPrice.asText() ;

  System.out.println( String.format("Name : %s Url : %s Price : %s", itemName, itemPrice, itemUrl));
  }
}

Then instead of just printing the results, we are going to put it in JSON, using Jackson library, to map items in JSON format.

First we need a POJO (plain old java object) to represent Items

Item.java

public class Item {  
    private String title ; 
    private BigDecimal price ;
    private String url ;
//getters and setters
}

Then add this to your pom.xml :

<dependency>  
  <groupId>com.fasterxml.jackson.core</groupId>
  <artifactId>jackson-databind</artifactId>
  <version>2.7.0</version>
</dependency>  

Now all we have to do is create an Item, set its attributes, and convert it to JSON string (or a file ...), and adapt the previous code a little bit :

for(HtmlElement htmlItem : items){  
  HtmlAnchor itemAnchor = ((HtmlAnchor) htmlItem.getFirstByXPath(".//span[@class='txt']/span[@class='pl']/a"));
  HtmlElement spanPrice = ((HtmlElement) htmlItem.getFirstByXPath(".//span[@class='txt']/span[@class='l2']/span[@class='price']")) ;

  // It is possible that an item doesn't have any price, we set the price to 0.0 in this case
  String itemPrice = spanPrice == null ? "0.0" : spanPrice.asText() ;

  Item item = new Item();

  item.setTitle(itemAnchor.asText());

  item.setUrl( baseUrl +
  itemAnchor.getHrefAttribute());

  item.setPrice(new BigDecimal(itemPrice.replace("$", "")));
  ObjectMapper mapper = new ObjectMapper();
  String jsonString = mapper.writeValueAsString(item) ;
  System.out.println(jsonString);
}

Go further

This example is not perfect, there are many things that can be improved :

  • Multi-city search
  • Handling pagination
  • Multi criteria search

You can find the code in this Github repo

This was my first blog post I hope you enjoyed it, feel free to give me any feedback in the comments.

Next time

In the next posts I will talk about :

  • How to handle authentication with you scrapers
  • Pagination
  • Captchas breaking
  • Running phantomJS from Java to handle heavy Javascript website
  • many more !

Kevin Sahin

You want to work with me for advices / custom scraping work ? Just send me an email at hi@ksah.in :)

Paris