ParseHtmlTableLocally.java

import com.gargoylesoftware.htmlunit.StringWebResponse;
import com.gargoylesoftware.htmlunit.WebClient;
import com.gargoylesoftware.htmlunit.html.DomElement;
import com.gargoylesoftware.htmlunit.html.HTMLParser;
import com.gargoylesoftware.htmlunit.html.HtmlPage;
import org.openqa.selenium.WebElement;
import org.testng.annotations.Test;

import java.io.IOException;
import java.net.URL;
import java.util.ArrayList;

public class ParseHtmlTableLocally extends SauceTestNGBase
{
	@Test
	void usingHtmlUnit() throws IOException
	{
		/* get table source HTML using Selenium */
		driver.get(PageWithTables.URL);
		WebElement element = driver.findElement(PageWithTables.table);
		String html = element.getAttribute("outerHTML");
		System.out.println(html);
		
		/* build HtmlUnit HtmlPage from string representation <table>...</table> */
		URL url = new URL("http://www.example.com");
		StringWebResponse response = new StringWebResponse(html, url);
		WebClient client = new WebClient();
		HtmlPage page = HTMLParser.parseHtml(response, client.getCurrentWindow());
		
		/* create object to save results */
		ArrayList<ArrayList<String>> data = new ArrayList<>();
		
		/* get headings */
		ArrayList<String> headings = new ArrayList<>();
		for (DomElement th : page.getElementsByTagName("th"))
		{
			String heading = th.getTextContent();
			headings.add(heading);
		}
		data.add(headings);
		
		/* get each row */
		for (DomElement tr : page.getElementsByTagName("tr"))
		{
			ArrayList<String> row = new ArrayList<>();
			
			/* get each cell from the row */
			for(DomElement td : tr.getElementsByTagName("td"))
			{
				String cell = td.getTextContent();
				row.add(cell);
			}
			
			if (row.size() > 0)
			{
				data.add(row);
			}
		
		}
		
		System.out.println("data" + data);
	}
}

SauceTestNGBase.java

import org.openqa.selenium.WebDriver;
import org.openqa.selenium.remote.DesiredCapabilities;
import org.openqa.selenium.remote.RemoteWebDriver;
import org.testng.annotations.AfterMethod;
import org.testng.annotations.BeforeMethod;

import java.net.MalformedURLException;
import java.net.URL;

public abstract class SauceTestNGBase
{
	public static final String SAUCE_URL = "https://SAUCE_USERNAME:[email protected]:443/wd/hub";
	public String SAUCE_USERNAME = System.getenv("SAUCE_USERNAME");
	public String SAUCE_ACCESS_KEY = System.getenv("SAUCE_ACCESS_KEY");
	
	public String SELENIUM_PLATFORM = "Windows 10";
	public String SELENIUM_BROWSER = "Firefox";
	public String SELENIUM_VERSION = "latest";
		
  public String testName = "Default Test Name";
  
	WebDriver driver;
	
	@BeforeMethod
	public void setup() throws MalformedURLException
	{
		URL remoteUrl = getSauceURL(SAUCE_USERNAME, SAUCE_ACCESS_KEY);
		DesiredCapabilities capabilities = getDesiredCapabilities(SELENIUM_PLATFORM, SELENIUM_BROWSER, SELENIUM_VERSION);
		driver = new RemoteWebDriver(remoteUrl, capabilities);
	}
	
	@AfterMethod
	public void teardown()
	{
		driver.quit();
	}
	
	
	public static URL getSauceURL(String username, String accessKey) throws MalformedURLException
	{
		URL url = new URL(SAUCE_URL
				.replace("SAUCE_USERNAME", username)
				.replace("SAUCE_ACCESS_KEY", accessKey));
		
		return url;
	}
	
	public DesiredCapabilities getDesiredCapabilities(String platform, String browserName, String browserVersion)
	{
		DesiredCapabilities capabilities = new DesiredCapabilities();
		capabilities.setCapability("platform", platform);
		capabilities.setCapability("browserName", browserName);
		capabilities.setCapability("version", browserVersion);
		capabilities.setCapability("name", testName);
		
		return capabilities;
	}
}