Update dependencies and fix issue with jsoup

An issue appeared when upgrading jsoup. They must have changed the
implementation of the "remove" method of "Elements.iterator()". Calling
this method now removes the HTML element from the document, instead of
just removing it from the Elements list.

Author: mangstadt

pom.xml

@@ -68,22 +68,22 @@
 		<dependency>
 			<groupId>org.jsoup</groupId>
 			<artifactId>jsoup</artifactId>
-			<version>1.16.1</version>
+			<version>1.22.1</version>
 		</dependency>
 		<dependency>
 			<groupId>org.freemarker</groupId>
 			<artifactId>freemarker</artifactId>
-			<version>2.3.32</version>
+			<version>2.3.34</version>
 		</dependency>
 		<dependency>
 			<groupId>com.fasterxml.jackson.core</groupId>
 			<artifactId>jackson-core</artifactId>
-			<version>2.15.2</version>
+			<version>2.21.0</version>
 		</dependency>
 		<dependency>
 			<groupId>com.fasterxml.jackson.core</groupId>
 			<artifactId>jackson-databind</artifactId>
-			<version>2.15.2</version>
+			<version>2.21.0</version>
 
 			<!--
 			The only classes that refer to objects from jackson-databind are JCardModule, JCardSerializer, and JCardDeserializer, all of which

src/main/java/ezvcard/io/html/HCardParser.java

@@ -6,9 +6,9 @@
 import java.io.InputStream;
 import java.io.Reader;
 import java.io.UncheckedIOException;
-import java.net.MalformedURLException;
 import java.net.URL;
 import java.nio.file.Path;
+import java.time.Duration;
 import java.util.ArrayList;
 import java.util.Iterator;
 import java.util.List;
@@ -39,6 +39,7 @@
 import ezvcard.property.Url;
 import ezvcard.property.VCardProperty;
 import ezvcard.util.Gobble;
+import ezvcard.util.HtmlUtils;
 import ezvcard.util.IOUtils;
 
 /*
@@ -88,8 +89,9 @@
  * wiki/hcard</a>
  */
 public class HCardParser extends StreamReader {
+	private static final Duration urlTimeout = Duration.ofSeconds(30);
+
 	private final String pageUrl;
-	private final Elements vcardElements;
 	private final Iterator<Element> vcardElementsIt;
 	private final List<Label> labels = new ArrayList<>();
 
@@ -109,7 +111,7 @@ public class HCardParser extends StreamReader {
 	 * @throws IOException if there's a problem loading the webpage
 	 */
 	public HCardParser(URL url) throws IOException {
-		this(Jsoup.parse(url, 30000), url.toString());
+		this(Jsoup.parse(url, (int) urlTimeout.toMillis()), url.toString());
 	}
 
 	/**
@@ -207,36 +209,30 @@ public HCardParser(Document document) {
 	public HCardParser(Document document, String pageUrl) {
 		this.pageUrl = pageUrl;
 
-		String anchor = null;
-		if (pageUrl != null) {
-			try {
-				URL url = new URL(pageUrl);
-				anchor = url.getRef();
-			} catch (MalformedURLException e) {
-				anchor = null;
-			}
-		}
+		String anchor = (pageUrl == null) ? null : HtmlUtils.getAnchorFromUrl(pageUrl);
 
-		Element searchUnder = null;
-		if (anchor != null) {
-			searchUnder = document.getElementById(anchor);
-		}
+		Element searchUnder = (anchor == null) ? null : document.getElementById(anchor);
 		if (searchUnder == null) {
 			searchUnder = document;
 		}
 
-		vcardElements = searchUnder.getElementsByClass("vcard");
+		/*
+		 * Nested vCards also show up in this list as separate list items. For
+		 * example, if the HTML document has one vCard and that vCard has one
+		 * nested vCard (i.e. AGENT property), this list will have two elements.
+		 * 
+		 * Exclude the nested vCards from being processed as their own,
+		 * independent vCards.
+		 */
+		Elements vcardElementsIncludingNested = searchUnder.getElementsByClass("vcard");
 
-		//remove nested vcard elements
-		Iterator<Element> it = vcardElements.iterator();
-		while (it.hasNext()) {
-			Element element = it.next();
-			if (isChildOf(element, vcardElements)) {
-				it.remove();
-			}
-		}
+		//@formatter:off
+		Elements vcardElementsWithoutNested = new Elements(vcardElementsIncludingNested.stream()
+			.filter(element -> !isChildOf(element, vcardElementsIncludingNested))
+		.collect(Collectors.toList()));
+		//@formatter:on
 
-		vcardElementsIt = vcardElements.iterator();
+		vcardElementsIt = vcardElementsWithoutNested.iterator();
 	}
 
 	/**
@@ -246,8 +242,7 @@ public HCardParser(Document document, String pageUrl) {
 	 */
 	private HCardParser(Element embeddedVCard, String pageUrl) {
 		this.pageUrl = pageUrl;
-		vcardElements = new Elements(embeddedVCard);
-		vcardElementsIt = vcardElements.iterator();
+		vcardElementsIt = new Elements(embeddedVCard).iterator();
 	}
 
 	@Override

src/main/java/ezvcard/util/HtmlUtils.java

@@ -1,5 +1,8 @@
 package ezvcard.util;
 
+import java.net.MalformedURLException;
+import java.net.URL;
+
 import org.jsoup.Jsoup;
 import org.jsoup.nodes.Document;
 import org.jsoup.nodes.Element;
@@ -70,6 +73,23 @@ public static Element toElement(String html, String baseUrl) {
 		return d.getElementsByTag("body").first().children().first();
 	}
 
+	/**
+	 * Extracts the anchor portion from a URL.
+	 * @param url the full URL
+	 * @return the anchor, or null if there is no anchor, or null if the given
+	 * string is not a valid URL
+	 */
+	public static String getAnchorFromUrl(String url) {
+		URL urlObj;
+		try {
+			urlObj = new URL(url);
+		} catch (MalformedURLException e) {
+			return null;
+		}
+
+		return urlObj.getRef();
+	}
+
 	private HtmlUtils() {
 		//hide
 	}

src/test/java/ezvcard/util/HtmlUtilsTest.java

@@ -2,6 +2,7 @@
 
 import static org.junit.Assert.assertEquals;
 import static org.junit.Assert.assertFalse;
+import static org.junit.Assert.assertNull;
 import static org.junit.Assert.assertTrue;
 
 import org.jsoup.Jsoup;
@@ -80,4 +81,12 @@ public void toElement_with_base_url() {
 		assertEquals("img", element.tagName());
 		assertEquals("http://example.com/image.png", element.absUrl("src"));
 	}
+
+	@Test
+	public void getAnchorFromUrl() {
+		assertEquals("foo", HtmlUtils.getAnchorFromUrl("https://domain.com/page#foo"));
+		assertEquals("", HtmlUtils.getAnchorFromUrl("https://domain.com/page#"));
+		assertNull(HtmlUtils.getAnchorFromUrl("https://domain.com/page"));
+		assertNull(HtmlUtils.getAnchorFromUrl("not a URL"));
+	}
 }