#53: Handle and extract text content from DOCX and ODT files

docs-core/src/test/java/com/sismics/docs/core/util/TestEncryptUtil.java

@@ -18,7 +18,6 @@ import com.google.common.io.ByteStreams;
  * @author bgamard
  */
 public class TestEncryptUtil {
-
     /**
      * Test private key.
      */

docs-core/src/test/java/com/sismics/docs/core/util/TestFileUtil.java

@@ -0,0 +1,36 @@
+package com.sismics.docs.core.util;
+
+import java.io.InputStream;
+
+import junit.framework.Assert;
+
+import org.junit.Test;
+
+import com.google.common.io.Resources;
+import com.sismics.docs.core.model.jpa.File;
+import com.sismics.util.mime.MimeType;
+
+/**
+ * Test of the file entity utilities.
+ * 
+ * @author bgamard
+ */
+public class TestFileUtil {
+    @Test
+    public void extractContentOpenDocumentTextTest() throws Exception {
+        try (InputStream inputStream = Resources.getResource("file/document.odt").openStream()) {
+            File file = new File();
+            file.setMimeType(MimeType.OPEN_DOCUMENT_TEXT);
+            Assert.assertEquals("Lorem ipsum dolor sit amen.\r\n", FileUtil.extractContent(null, file, inputStream));
+        }
+    }
+    
+    @Test
+    public void extractContentOfficeDocumentTest() throws Exception {
+        try (InputStream inputStream = Resources.getResource("file/document.docx").openStream()) {
+            File file = new File();
+            file.setMimeType(MimeType.OFFICE_DOCUMENT);
+            Assert.assertEquals("Lorem ipsum dolor sit amen.\r\n", FileUtil.extractContent(null, file, inputStream));
+        }
+    }
+}

docs-core/src/test/java/com/sismics/util/TestMimeTypeUtil.java

@@ -0,0 +1,40 @@
+package com.sismics.util;
+
+import java.io.ByteArrayInputStream;
+import java.io.InputStream;
+
+import org.apache.commons.compress.utils.IOUtils;
+import org.junit.Assert;
+import org.junit.Test;
+
+import com.google.common.io.Resources;
+import com.sismics.docs.core.model.jpa.File;
+import com.sismics.util.mime.MimeType;
+import com.sismics.util.mime.MimeTypeUtil;
+
+/**
+ * Test of the utilities to check MIME types.
+ * 
+ * @author bgamard
+ */
+public class TestMimeTypeUtil {
+
+    @Test
+    public void guessOpenDocumentFormatTest() throws Exception {
+        // Detect ODT files
+        try (InputStream inputStream = Resources.getResource("file/document.odt").openStream();
+                InputStream byteArrayInputStream = new ByteArrayInputStream(IOUtils.toByteArray(inputStream))) {
+            File file = new File();
+            file.setMimeType(MimeType.APPLICATION_ZIP);
+            Assert.assertEquals(MimeType.OPEN_DOCUMENT_TEXT, MimeTypeUtil.guessOpenDocumentFormat(file, byteArrayInputStream));
+        }
+        
+        // Detect DOCX files
+        try (InputStream inputStream = Resources.getResource("file/document.docx").openStream();
+                InputStream byteArrayInputStream = new ByteArrayInputStream(IOUtils.toByteArray(inputStream))) {
+            File file = new File();
+            file.setMimeType(MimeType.APPLICATION_ZIP);
+            Assert.assertEquals(MimeType.OFFICE_DOCUMENT, MimeTypeUtil.guessOpenDocumentFormat(file, byteArrayInputStream));
+        }
+    }
+}