Merge pull request #741 from emmanuel-keller/1.6

Fix #739. HTMLCleaner does not preserve the lang attribute on DOM
jaeksoft · Jul 1, 2014 · 4170d57 · 4170d57
2 parents 864b82f + 00ba281
commit 4170d57
Showing 1 changed file with 3 additions and 0 deletions.
diff --git a/src/main/java/com/jaeksoft/searchlib/parser/htmlParser/HtmlCleanerParser.java b/src/main/java/com/jaeksoft/searchlib/parser/htmlParser/HtmlCleanerParser.java
@@ -80,6 +80,9 @@ protected HtmlNodeAbstract<?> getDocument(String pageSource)
 	private DomHtmlNode getDomHtmlNode() throws ParserConfigurationException {
 		Document document = new DomSerializer(cleaner.getProperties(), true)
 				.createDOM(rootTagNode);
+		String lang = rootTagNode.getAttributeByName("lang");
+		if (lang != null)
+			document.getDocumentElement().setAttribute("lang", lang);
 		return new DomHtmlNode(document);
 	}