Use jsoup in HtmlProcessor and HtmlSanitizer (268189c1) · Commits · e / os / Mail

k9mail/build.gradle

+2 −1

Original line number	Diff line number	Diff line
		@@ -28,6 +28,7 @@ dependencies {
		compile 'commons-io:commons-io:2.4'
		compile "com.android.support:support-v4:${androidSupportLibraryVersion}"
		compile 'net.sourceforge.htmlcleaner:htmlcleaner:2.18'
		compile 'org.jsoup:jsoup:1.10.2'
		compile 'de.cketti.library.changelog:ckchangelog:1.2.1'
		compile 'com.github.bumptech.glide:glide:3.6.1'
		compile 'com.splitwise:tokenautocomplete:2.0.7'
		@@ -41,7 +42,6 @@ dependencies {
		testCompile "org.robolectric:robolectric:${robolectricVersion}"
		testCompile "junit:junit:${junitVersion}"
		testCompile "org.mockito:mockito-core:${mockitoVersion}"
		testCompile 'org.jsoup:jsoup:1.10.2'
		}

		android {
		@@ -96,6 +96,7 @@ android {
		exclude 'META-INF/LICENSE.txt'
		exclude 'META-INF/NOTICE'
		exclude 'META-INF/NOTICE.txt'
		exclude 'META-INF/README'
		exclude 'LICENSE.txt'
		}

k9mail/src/main/java/com/fsck/k9/message/html/HeadCleaner.java

0 → 100644

+100 −0

Original line number	Diff line number	Diff line
		package com.fsck.k9.message.html;


		import java.util.List;
		import java.util.Locale;

		import org.jsoup.nodes.Attributes;
		import org.jsoup.nodes.DataNode;
		import org.jsoup.nodes.Document;
		import org.jsoup.nodes.Element;
		import org.jsoup.nodes.Node;
		import org.jsoup.nodes.TextNode;
		import org.jsoup.parser.Tag;
		import org.jsoup.select.NodeTraversor;
		import org.jsoup.select.NodeVisitor;

		import static java.util.Arrays.asList;


		class HeadCleaner {
		private static final List<String> ALLOWED_TAGS = asList("style", "meta");


		public void clean(Document dirtyDocument, Document cleanedDocument) {
		copySafeNodes(dirtyDocument.head(), cleanedDocument.head());
		}

		private void copySafeNodes(Element source, Element destination) {
		CleaningVisitor cleaningVisitor = new CleaningVisitor(source, destination);
		NodeTraversor traversor = new NodeTraversor(cleaningVisitor);
		traversor.traverse(source);
		}


		static class CleaningVisitor implements NodeVisitor {
		private final Element root;
		private Element destination;
		private boolean skipChildren = false;


		CleaningVisitor(Element root, Element destination) {
		this.root = root;
		this.destination = destination;
		}

		public void head(Node source, int depth) {
		if (skipChildren) {
		return;
		}

		if (source instanceof Element) {
		Element sourceElement = (Element) source;

		if (isSafeTag(sourceElement)) {
		String sourceTag = sourceElement.tagName();
		Attributes destinationAttributes = sourceElement.attributes().clone();
		Element destinationChild = new Element(Tag.valueOf(sourceTag), sourceElement.baseUri(), destinationAttributes);

		destination.appendChild(destinationChild);
		destination = destinationChild;
		} else if (source != root) {
		skipChildren = true;
		}
		} else if (source instanceof TextNode) {
		TextNode sourceText = (TextNode) source;
		TextNode destinationText = new TextNode(sourceText.getWholeText(), source.baseUri());
		destination.appendChild(destinationText);
		} else if (source instanceof DataNode && isSafeTag(source.parent())) {
		DataNode sourceData = (DataNode) source;
		DataNode destinationData = new DataNode(sourceData.getWholeData(), source.baseUri());
		destination.appendChild(destinationData);
		}
		}

		public void tail(Node source, int depth) {
		if (source == destination) {
		destination = destination.parent();
		skipChildren = false;
		}
		}

		private boolean isSafeTag(Node node) {
		if (isMetaRefresh(node)) {
		return false;
		}

		String tag = node.nodeName().toLowerCase(Locale.ROOT);
		return ALLOWED_TAGS.contains(tag);
		}

		private boolean isMetaRefresh(Node node) {
		if (!"meta".equalsIgnoreCase(node.nodeName())) {
		return false;
		}

		String attributeValue = node.attributes().getIgnoreCase("http-equiv");
		return "refresh".equalsIgnoreCase(attributeValue.trim());
		}
		}
		}

k9mail/src/main/java/com/fsck/k9/message/html/HtmlConverter.java

+2 −2

Original line number	Diff line number	Diff line
		@@ -1260,7 +1260,7 @@ public class HtmlConverter {
		"</body></html>";
		}

		private static String cssStyleTheme() {
		static String cssStyleTheme() {
		if (K9.getK9MessageViewTheme() == K9.Theme.DARK) {
		return "<style type=\"text/css\">" +
		"* { background: black ! important; color: #F3F3F3 !important }" +
		@@ -1283,7 +1283,7 @@ public class HtmlConverter {
		* A {@code <style>} element that can be dynamically included in the HTML
		* {@code <head>} element when messages are displayed.
		*/
		private static String cssStylePre() {
		static String cssStylePre() {
		final String font = K9.messageViewFixedWidthFont()
		? "monospace"
		: "sans-serif";

k9mail/src/main/java/com/fsck/k9/message/html/HtmlProcessor.java

+22 −3

Original line number	Diff line number	Diff line
		package com.fsck.k9.message.html;


		import org.jsoup.nodes.Document;


		public class HtmlProcessor {
		private final HtmlSanitizer htmlSanitizer;


		public static HtmlProcessor newInstance() {
		HtmlSanitizer htmlSanitizer = HtmlSanitizer.getInstance();
		HtmlSanitizer htmlSanitizer = new HtmlSanitizer();
		return new HtmlProcessor(htmlSanitizer);
		}

		@@ -15,7 +18,23 @@ public class HtmlProcessor {
		}

		public String processForDisplay(String html) {
		String wrappedHtml = HtmlConverter.wrapMessageContent(html);
		return htmlSanitizer.sanitize(wrappedHtml);
		Document document = htmlSanitizer.sanitize(html);
		addCustomHeadContents(document);

		return toCompactString(document);
		}

		private void addCustomHeadContents(Document document) {
		document.head().append("<meta name=\"viewport\" content=\"width=device-width\"/>" +
		HtmlConverter.cssStyleTheme() +
		HtmlConverter.cssStylePre());
		}

		static String toCompactString(Document document) {
		document.outputSettings()
		.prettyPrint(false)
		.indentAmount(0);

		return document.html();
		}
		}

k9mail/src/main/java/com/fsck/k9/message/html/HtmlSanitizer.java

+21 −53

Original line number	Diff line number	Diff line
		package com.fsck.k9.message.html;


		import android.support.annotation.VisibleForTesting;

		import org.htmlcleaner.CleanerProperties;
		import org.htmlcleaner.HtmlCleaner;
		import org.htmlcleaner.HtmlSerializer;
		import org.htmlcleaner.SimpleHtmlSerializer;
		import org.htmlcleaner.TagNode;
		import org.jsoup.Jsoup;
		import org.jsoup.nodes.Document;
		import org.jsoup.safety.Cleaner;
		import org.jsoup.safety.Whitelist;


		public class HtmlSanitizer {
		private static final HtmlCleaner HTML_CLEANER;
		private static final HtmlSerializer HTML_SERIALIZER;

		static {
		CleanerProperties properties = createCleanerProperties();
		HTML_CLEANER = new HtmlCleaner(properties);
		HTML_SERIALIZER = new SimpleHtmlSerializer(properties);
		}
		private final HeadCleaner headCleaner;
		private final Cleaner cleaner;

		HtmlSanitizer() {
		Whitelist whitelist = Whitelist.relaxed()
		.addTags("font")
		.addAttributes("table", "align", "bgcolor", "border", "cellpadding", "cellspacing", "width")
		.addAttributes(":all", "class", "style", "id")
		.addProtocols("img", "src", "http", "https", "cid", "data");

		public static HtmlSanitizer getInstance() {
		return new HtmlSanitizer();
		cleaner = new Cleaner(whitelist);
		headCleaner = new HeadCleaner();
		}

		@VisibleForTesting
		HtmlSanitizer() {}


		public String sanitize(String html) {
		TagNode rootNode = HTML_CLEANER.clean(html);

		removeMetaRefresh(rootNode);

		return HTML_SERIALIZER.getAsString(rootNode, "UTF8");
		}

		private static CleanerProperties createCleanerProperties() {
		CleanerProperties properties = new CleanerProperties();

		// See http://htmlcleaner.sourceforge.net/parameters.php for descriptions
		properties.setNamespacesAware(false);
		properties.setAdvancedXmlEscape(false);
		properties.setOmitXmlDeclaration(true);
		properties.setOmitDoctypeDeclaration(false);
		properties.setTranslateSpecialEntities(false);
		properties.setRecognizeUnicodeChars(false);
		properties.setIgnoreQuestAndExclam(false);

		return properties;
		}

		private void removeMetaRefresh(TagNode rootNode) {
		for (TagNode element : rootNode.getElementListByName("meta", true)) {
		String httpEquiv = element.getAttributeByName("http-equiv");
		if (httpEquiv != null && httpEquiv.trim().equalsIgnoreCase("refresh")) {
		element.removeFromTree();
		}
		}
		public Document sanitize(String html) {
		Document dirtyDocument = Jsoup.parse(html);
		Document cleanedDocument = cleaner.clean(dirtyDocument);
		headCleaner.clean(dirtyDocument, cleanedDocument);
		return cleanedDocument;
		}
		}