Add extract_text_only option

src/com/dom_distiller/client/ContentExtractor.java

Index: src/com/dom_distiller/client/ContentExtractor.java
diff --git a/src/com/dom_distiller/client/ContentExtractor.java b/src/com/dom_distiller/client/ContentExtractor.java
index 2491fb616b1010050894261f6eaf5bbc39fb53bc..50cfef361ea98e891243907e6e5f7b2782d4eead 100644
--- a/src/com/dom_distiller/client/ContentExtractor.java
+++ b/src/com/dom_distiller/client/ContentExtractor.java
@@ -35,6 +35,10 @@ public class ContentExtractor implements Exportable {
     static Logger logger = Logger.getLogger("DomDistiller");
 
     public static String extractContent() {
+        return extractContent(false);
+    }
+
+    public static String extractContent(boolean text_only) {
         BoilerpipeHTMLContentHandler htmlParser = new BoilerpipeHTMLContentHandler();
         List<Node> textNodes = null;
 
@@ -56,19 +60,11 @@ public class ContentExtractor implements Exportable {
             return "";
         }
 
-        List<Integer> contentTextIndexes = new ArrayList<Integer>();
-        for (TextBlock tb : document.getTextBlocks()) {
-            if (!tb.hasLabel(DefaultLabels.TITLE)) {
-                contentTextIndexes.addAll(tb.getContainedTextElements());
-            }
+        if (text_only) {
+            return document.getText(true, false);
         }
-        Collections.sort(contentTextIndexes);
 
-        // Boilerpipe's text node indexes start at 1.
-        List<Node> contentNodes = new ArrayList<Node>(contentTextIndexes.size());
-        for (Integer i : contentTextIndexes) {
-            contentNodes.add(textNodes.get(i - 1));
-        }
+        List<Node> contentNodes = getContentNodesForTextDocument(document, textNodes);
 
         List<Node> contentAndImages = RelevantImageFinder.findAndAddImages(
                 contentNodes, Document.get().getDocumentElement());
@@ -100,6 +96,24 @@ public class ContentExtractor implements Exportable {
         return domToSaxVisitor.getTextNodes();
     }
 
+    private static List<Node> getContentNodesForTextDocument(
+            TextDocument document, List<Node> textNodes) {
+        List<Integer> contentTextIndexes = new ArrayList<Integer>();
+        for (TextBlock tb : document.getTextBlocks()) {
+            if (!tb.hasLabel(DefaultLabels.TITLE)) {
+                contentTextIndexes.addAll(tb.getContainedTextElements());
+            }
+        }
+        Collections.sort(contentTextIndexes);
+
+        // Boilerpipe's text node indexes start at 1.
+        List<Node> contentNodes = new ArrayList<Node>(contentTextIndexes.size());
+        for (Integer i : contentTextIndexes) {
+            contentNodes.add(textNodes.get(i - 1));
+        }
+        return contentNodes;
+    }
+
     private static void makeAllLinksAbsolute(Node rootNode) {
         Element root = Element.as(rootNode);