Fixed TextMining (Word to Text) transformer that was not closing an input stream after use

Added checks to the Abstract transformer to log an error non-closures of streams.  This won't halt execution.


git-svn-id: https://svn.alfresco.com/repos/alfresco-enterprise/alfresco/HEAD/root@2384 c4b6b30b-aa2e-2d43-bbcb-ca4b014f7261

source/java/org/alfresco/repo/content/AbstractContentAccessor.java

@@ -88,7 +88,7 @@ public abstract class AbstractContentAccessor implements ContentAccessor
             {
                 StringBuilder sb = new StringBuilder(1024);
                 StackTraceUtil.buildStackTrace(
-                        "Content IO Channel was opened but not closed",
+                        "Content IO Channel was opened but not closed: \n" + this,
                         traceLoggerChannelAssignTrace,
                         sb,
                         -1);

source/java/org/alfresco/repo/content/transform/AbstractContentTransformer.java

@@ -240,6 +240,18 @@ public abstract class AbstractContentTransformer implements ContentTransformer
                     "   options: " + options,
                     e);
         }
+        finally
+        {
+            // check that the reader and writer are both closed
+            if (!reader.isClosed())
+            {
+                logger.error("Content reader not closed by transformer: \n" + reader);
+            }
+            if (!writer.isClosed())
+            {
+                logger.error("Content writer not closed by transformer: \n" + writer);
+            }
+        }
         
         // record time
         long after = System.currentTimeMillis();

source/java/org/alfresco/repo/content/transform/ContentTransformer.java

@@ -70,6 +70,8 @@ public interface ContentTransformer
      * The source and target mimetypes <b>must</b> be available on the
      * {@link org.alfresco.service.cmr.repository.ContentAccessor#getMimetype()} methods of
      * both the reader and the writer.
+     * <p>
+     * Both reader and writer will be closed after the transformation completes.
      * 
      * @param reader the source of the content
      * @param writer the destination of the transformed content

source/java/org/alfresco/repo/content/transform/TextMiningContentTransformer.java

@@ -23,8 +23,6 @@ import java.util.Map;
 import org.alfresco.repo.content.MimetypeMap;
 import org.alfresco.service.cmr.repository.ContentReader;
 import org.alfresco.service.cmr.repository.ContentWriter;
-import org.apache.commons.logging.Log;
-import org.apache.commons.logging.LogFactory;
 import org.textmining.text.extraction.WordExtractor;
 
 /**
@@ -35,8 +33,6 @@ import org.textmining.text.extraction.WordExtractor;
  */
 public class TextMiningContentTransformer extends AbstractContentTransformer
 {
-    private static final Log logger = LogFactory.getLog(TextMiningContentTransformer.class);
-    
     private WordExtractor wordExtractor;
     
     public TextMiningContentTransformer()
@@ -64,10 +60,11 @@ public class TextMiningContentTransformer extends AbstractContentTransformer
     public void transformInternal(ContentReader reader, ContentWriter writer,  Map<String, Object> options)
             throws Exception
     {
-        InputStream is = reader.getContentInputStream();
+        InputStream is = null;
         String text = null;
         try
         {
+            is = reader.getContentInputStream();
             text = wordExtractor.extractText(is);
         }
         catch (IOException e)
@@ -80,7 +77,14 @@ public class TextMiningContentTransformer extends AbstractContentTransformer
                 text = "";
             }
         }
-        // dump the text out
+        finally
+        {
+            if (is != null)
+            {
+                is.close();
+            }
+        }
+        // dump the text out.  This will close the writer automatically.
         writer.putContent(text);
     }
 }