Daenyth/ImageExtractor.java

## ImageExtractor.java
import java.io.IOException;
import java.net.URL;

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

/**
 * Given a url to a web page, extract a suitable image from that page. This will
 * attempt to follow a method similar to Google+, as described <a href=
 * "http://webmasters.stackexchange.com/questions/25581/how-does-google-plus-select-an-image-from-a-shared-link"
 * >here</a>
 *
 */
public class ImageExtractor {

    // TODO: Add junit test case for this. (Construct Document from string, extract, check)
    public static String extractImageUrl(String url) throws IOException {
        String contentType = new URL(url).openConnection().getContentType();
        if (contentType != null) {
            if (contentType.startsWith("image/")) {
                return url;
            }
        }

        Document document = Jsoup.connect(url).get();

        String imageUrl = null;

        imageUrl = getImageFromSchema(document);
        if (imageUrl != null) {
            return imageUrl;
        }

        imageUrl = getImageFromOpenGraph(document);
        if (imageUrl != null) {
            return imageUrl;
        }

        imageUrl = getImageFromTwitterCard(document);
        if (imageUrl != null) {
            return imageUrl;
        }

        imageUrl = getImageFromTwitterShared(document);
        if (imageUrl != null) {
            return imageUrl;
        }

        imageUrl = getImageFromLinkRel(document);
        if (imageUrl != null) {
            return imageUrl;
        }

        imageUrl = getImageFromGuess(document);
        if (imageUrl != null) {
            return imageUrl;
        }

        return imageUrl;
    }

    private static String getImageFromTwitterShared(Document document) {
        Element div = document.select("div.media-gallery-image-wrapper").first();
        if (div == null) {
            return null;
        }
        Element img = div.select("img.media-slideshow-image").first();
        if (img != null) {
            return img.absUrl("src");
        }
        return null;
    }

    private static String getImageFromGuess(Document document) {
        // TODO
        return null;
    }

    private static String getImageFromLinkRel(Document document) {
        Element link = document.select("link[rel=image_src]").first();
        if (link != null) {
            return link.attr("abs:href");
        }
        return null;
    }

    private static String getImageFromTwitterCard(Document document) {
        Element meta = document.select("meta[name=twitter:card][content=photo]").first();
        if (meta == null) {
            return null;
        }
        Element image = document.select("meta[name=twitter:image]").first();
        return image.attr("abs:content");
    }

    private static String getImageFromOpenGraph(Document document) {
        Element image = document.select("meta[property=og:image]").first();
        if (image != null) {
            return image.attr("abs:content");
        }
        Element secureImage = document.select("meta[property=og:image:secure]").first();
        if (secureImage != null) {
            return secureImage.attr("abs:content");
        }
        return null;
    }

    private static String getImageFromSchema(Document document) {
        Element container =
            document.select("*[itemscope][itemtype=http://schema.org/ImageObject]").first();
        if (container == null) {
            return null;
        }

        Element image = container.select("img[itemprop=contentUrl]").first();
        if (image == null) {
            return null;
        }
        return image.absUrl("src");
    }
}
	import java.io.IOException;
	import java.net.URL;

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

	/**
	* Given a url to a web page, extract a suitable image from that page. This will
	* attempt to follow a method similar to Google+, as described <a href=
	* "http://webmasters.stackexchange.com/questions/25581/how-does-google-plus-select-an-image-from-a-shared-link"
	* >here</a>
	*
	*/
	public class ImageExtractor {

	// TODO: Add junit test case for this. (Construct Document from string, extract, check)
	public static String extractImageUrl(String url) throws IOException {
	String contentType = new URL(url).openConnection().getContentType();
	if (contentType != null) {
	if (contentType.startsWith("image/")) {
	return url;
	}
	}

	Document document = Jsoup.connect(url).get();

	String imageUrl = null;

	imageUrl = getImageFromSchema(document);
	if (imageUrl != null) {
	return imageUrl;
	}

	imageUrl = getImageFromOpenGraph(document);
	if (imageUrl != null) {
	return imageUrl;
	}

	imageUrl = getImageFromTwitterCard(document);
	if (imageUrl != null) {
	return imageUrl;
	}

	imageUrl = getImageFromTwitterShared(document);
	if (imageUrl != null) {
	return imageUrl;
	}

	imageUrl = getImageFromLinkRel(document);
	if (imageUrl != null) {
	return imageUrl;
	}

	imageUrl = getImageFromGuess(document);
	if (imageUrl != null) {
	return imageUrl;
	}

	return imageUrl;
	}

	private static String getImageFromTwitterShared(Document document) {
	Element div = document.select("div.media-gallery-image-wrapper").first();
	if (div == null) {
	return null;
	}
	Element img = div.select("img.media-slideshow-image").first();
	if (img != null) {
	return img.absUrl("src");
	}
	return null;
	}

	private static String getImageFromGuess(Document document) {
	// TODO
	return null;
	}

	private static String getImageFromLinkRel(Document document) {
	Element link = document.select("link[rel=image_src]").first();
	if (link != null) {
	return link.attr("abs:href");
	}
	return null;
	}

	private static String getImageFromTwitterCard(Document document) {
	Element meta = document.select("meta[name=twitter:card][content=photo]").first();
	if (meta == null) {
	return null;
	}
	Element image = document.select("meta[name=twitter:image]").first();
	return image.attr("abs:content");
	}

	private static String getImageFromOpenGraph(Document document) {
	Element image = document.select("meta[property=og:image]").first();
	if (image != null) {
	return image.attr("abs:content");
	}
	Element secureImage = document.select("meta[property=og:image:secure]").first();
	if (secureImage != null) {
	return secureImage.attr("abs:content");
	}
	return null;
	}

	private static String getImageFromSchema(Document document) {
	Element container =
	document.select("*[itemscope][itemtype=http://schema.org/ImageObject]").first();
	if (container == null) {
	return null;
	}

	Element image = container.select("img[itemprop=contentUrl]").first();
	if (image == null) {
	return null;
	}
	return image.absUrl("src");
	}
	}