Scraper

package com.anemortalid.essex.whatson.scrape;

import java.util.ArrayList;
import java.util.List;

import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import org.jsoup.select.Elements;

/**
 * Scrapes the domain for some info
 *
 */
public class Scraper {

	private static final String ESSEX_DOMAIN = "http://www.essexstudent.com";

	public static void main(String[] args) {
		List<EssexEvent> events = new Scraper().scrubWhatson();
		System.out.println("There are " + events.size() + " events.");
		for (EssexEvent essexEvent : events) {
			System.out.println(essexEvent);
		}
	}

	public List<EssexEvent> scrubWhatson() {
		List<EssexEvent> events = new ArrayList<Scraper.EssexEvent>();

		try {
			Document document = Jsoup.connect(ESSEX_DOMAIN + "/whatson/").get();

			// Get events only, they most likely all have this class
			Elements eventItems = document.getElementsByClass("event_item");

			// extract the info, we don't really need to visit their links
			for (Element eventItem : eventItems) {

				EssexEvent event = new EssexEvent();

				Elements eventNames = eventItem.select("a.msl_event_name");
				Element aTag = eventNames.get(0);
				String link = aTag.attr("href");
				String title = aTag.text();
				event.setTitle(title);
				event.setEventLink(link);

				// These other tags have some more info, they're in a
				// div->dl->dd section
				Elements eventTimes = eventItem.getElementsByClass("msl_event_time");
				String eventTime = eventTimes.text();
				event.setTime(eventTime);

				Elements eventLocations = eventItem.getElementsByClass("msl_event_location");
				String eventLocation = eventLocations.text();
				event.setLocation(eventLocation);

				Elements eventDescriptions = eventItem.getElementsByClass("msl_event_description");
				String eventDescription = eventDescriptions.text();
				event.setDescription(eventDescription);

				// find the image at the end, not all event items have a
				// span>msl_event_image
				Elements eventImageElems = eventItem.select("span.msl_event_image");
				if (eventImageElems.size() > 0) {
					Element spanTag = eventImageElems.first();
					Elements spanChildren = spanTag.children();
					Element imgTag = spanChildren.first();
					String imgSource = imgTag.attr("src");
					event.setImgSrc(imgSource);
				}

				events.add(event);
			}

		} catch (Exception e) {
			// yolo
			e.printStackTrace();
		}

		return events;
	}

	/**
	 * Represents an Essex event
	 *
	 */
	private class EssexEvent {
		private String title;
		private String eventLink;
		private String imgSrc;
		private String time;
		private String location;
		private String description;

		public String getTitle() {
			return title;
		}

		public void setTitle(String title) {
			this.title = title;
		}

		public String getEventLink() {
			return eventLink;
		}

		public void setEventLink(String eventLink) {
			this.eventLink = eventLink;
		}

		public String getImgSrc() {
			return imgSrc;
		}

		public void setImgSrc(String imgSrc) {
			this.imgSrc = imgSrc;
		}

		public String getTime() {
			return time;
		}

		public void setTime(String time) {
			this.time = time;
		}

		public String getLocation() {
			return location;
		}

		public void setLocation(String location) {
			this.location = location;
		}

		public String getDescription() {
			return description;
		}

		public void setDescription(String description) {
			this.description = description;
		}

		public String getImageLocationLink() {
			if (imgSrc == null) {
				return "NO_IMAGE";
			}

			int questionMark = imgSrc.indexOf("?");
			return ESSEX_DOMAIN + imgSrc.substring(0, questionMark);
		}

		@Override
		public String toString() {
			StringBuilder builder = new StringBuilder();
			builder.append("EssexEvent [title=");
			builder.append(title);
			builder.append(", eventLink=");
			builder.append(eventLink);
			builder.append(", imgSrc=");
			builder.append(imgSrc);
			builder.append(", imgOnlyLink=");
			builder.append(getImageLocationLink());
			builder.append(", time=");
			builder.append(time);
			builder.append(", location=");
			builder.append(location);
			builder.append(", description=");
			builder.append(description);
			builder.append("]");
			return builder.toString();
		}
	}

}