SOLR-1274: added extract only output options

git-svn-id: https://svn.apache.org/repos/asf/lucene/solr/trunk@802282 13f79535-47bb-0310-9956-ffa450edef68

contrib/extraction/CHANGES.txt

@@ -34,3 +34,5 @@ $Id:$
     See http://www.lucidimagination.com/search/document/d6f1899a85b2a45c/vote_apache_tika_0_4_release_candidate_2#d6f1899a85b2a45c
     for discussion on language detection.
     See http://www.apache.org/dist/lucene/tika/CHANGES-0.4.txt. (gsingers)
+
+6. SOLR-1274: Added text serialization output for extractOnly (Peter Wolanin, gsingers)    

contrib/extraction/src/main/java/org/apache/solr/handler/extraction/ExtractingDocumentLoader.java

@@ -38,7 +38,9 @@ import org.apache.tika.sax.xpath.MatchingContentHandler;
 import org.apache.tika.sax.xpath.XPathParser;
 import org.apache.tika.exception.TikaException;
 import org.apache.xml.serialize.OutputFormat;
+import org.apache.xml.serialize.BaseMarkupSerializer;
 import org.apache.xml.serialize.XMLSerializer;
+import org.apache.xml.serialize.TextSerializer;
 import org.xml.sax.ContentHandler;
 import org.xml.sax.SAXException;
 
@@ -52,7 +54,14 @@ import java.io.StringWriter;
  *
  **/
 public class ExtractingDocumentLoader extends ContentStreamLoader {
-
+  /**
+   * Extract Only supported format
+   */
+  public static final String TEXT_FORMAT = "text";
+  /**
+   * Extract Only supported format.  Default
+   */
+  public static final String XML_FORMAT = "xml";
   /**
    * XHTML XPath parser.
    */
@@ -152,10 +161,17 @@ public class ExtractingDocumentLoader extends ContentStreamLoader {
         ContentHandler parsingHandler = handler;
 
         StringWriter writer = null;
-        XMLSerializer serializer = null;
+        BaseMarkupSerializer serializer = null;
         if (extractOnly == true) {
+          String extractFormat = params.get(ExtractingParams.EXTRACT_FORMAT, "xml");
           writer = new StringWriter();
+          if (extractFormat.equals(TEXT_FORMAT)) {
+            serializer = new TextSerializer();
+            serializer.setOutputCharStream(writer);
+            serializer.setOutputFormat(new OutputFormat("Text", "UTF-8", true));
+          } else {
             serializer = new XMLSerializer(writer, new OutputFormat("XML", "UTF-8", true));
+          }
           if (xpathExpr != null) {
             Matcher matcher =
                     PARSER.parse(xpathExpr);

contrib/extraction/src/main/java/org/apache/solr/handler/extraction/ExtractingParams.java

@@ -81,6 +81,11 @@ public interface ExtractingParams {
    */
   public static final String EXTRACT_ONLY = "extractOnly";
 
+  /**
+   * Content output format if extractOnly is true. Default is "xml", alternative is "text".
+   */
+  public static final String EXTRACT_FORMAT = "extractFormat";
+
   /**
    * Capture attributes separately according to the name of the element, instead of just adding them to the string buffer
    */

contrib/extraction/src/test/java/org/apache/solr/handler/ExtractingRequestHandlerTest.java

@@ -25,6 +25,7 @@ import org.apache.solr.common.util.NamedList;
 import org.apache.solr.common.SolrException;
 import org.apache.solr.handler.extraction.ExtractingParams;
 import org.apache.solr.handler.extraction.ExtractingRequestHandler;
+import org.apache.solr.handler.extraction.ExtractingDocumentLoader;
 
 import java.util.List;
 import java.util.ArrayList;
@@ -243,6 +244,24 @@ public class ExtractingRequestHandlerTest extends AbstractSolrTestCase {
     assertTrue("nl is null and it shouldn't be", nl != null);
     Object title = nl.get("title");
     assertTrue("title is null and it shouldn't be", title != null);
+    assertTrue(extraction.indexOf("<?xml") != -1);
+
+    rsp = loadLocal("solr-word.pdf", ExtractingParams.EXTRACT_ONLY, "true",
+            ExtractingParams.EXTRACT_FORMAT, ExtractingDocumentLoader.TEXT_FORMAT);
+    assertTrue("rsp is null and it shouldn't be", rsp != null);
+    list = rsp.getValues();
+
+    extraction = (String) list.get("solr-word.pdf");
+    assertTrue("extraction is null and it shouldn't be", extraction != null);
+    assertTrue(extraction + " does not contain " + "solr-word", extraction.indexOf("solr-word") != -1);
+    assertTrue(extraction.indexOf("<?xml") == -1);
+
+    nl = (NamedList) list.get("solr-word.pdf_metadata");
+    assertTrue("nl is null and it shouldn't be", nl != null);
+    title = nl.get("title");
+    assertTrue("title is null and it shouldn't be", title != null);
+
+
 
   }