More XSLF tests for the less common extensions, and initial support for .thmx (theme) files - currently just stretches to not breaking...

git-svn-id: https://svn.apache.org/repos/asf/poi/trunk@958923 13f79535-47bb-0310-9956-ffa450edef68

src/documentation/content/xdocs/status.xml

@@ -34,6 +34,7 @@
 
     <changes>
         <release version="3.7-beta2" date="2010-??-??">
+           <action dev="POI-DEVELOPERS" type="fix">XSLFSlideShow shouldn't break on .thmx (theme) files. Support for them is still very limited though</action>
         </release>
         <release version="3.7-beta1" date="2010-06-20">
            <action dev="POI-DEVELOPERS" type="fix">49432 - Lazy caching of XSSFComment CTComment objects by reference, to make repeated comment searching faster</action>

src/ooxml/java/org/apache/poi/POIXMLDocument.java

@@ -16,23 +16,30 @@
 ==================================================================== */
 package org.apache.poi;
 
-import java.io.*;
+import java.io.IOException;
-import java.util.*;
+import java.io.InputStream;
+import java.io.OutputStream;
+import java.io.PushbackInputStream;
+import java.util.HashMap;
+import java.util.HashSet;
+import java.util.List;
+import java.util.Map;
+import java.util.Set;
 
-import org.apache.poi.poifs.common.POIFSConstants;
-import org.apache.poi.util.IOUtils;
-import org.apache.poi.util.PackageHelper;
 import org.apache.poi.openxml4j.exceptions.InvalidFormatException;
 import org.apache.poi.openxml4j.exceptions.OpenXML4JException;
-import org.apache.poi.openxml4j.opc.*;
+import org.apache.poi.openxml4j.opc.OPCPackage;
+import org.apache.poi.openxml4j.opc.PackagePart;
+import org.apache.poi.openxml4j.opc.PackagePartName;
+import org.apache.poi.openxml4j.opc.PackageRelationship;
+import org.apache.poi.openxml4j.opc.PackageRelationshipCollection;
+import org.apache.poi.openxml4j.opc.PackagingURIHelper;
+import org.apache.poi.poifs.common.POIFSConstants;
+import org.apache.poi.util.IOUtils;
 
 public abstract class POIXMLDocument extends POIXMLDocumentPart{
     public static final String DOCUMENT_CREATOR = "Apache POI";
 
-    public static final String CORE_PROPERTIES_REL_TYPE = "http://schemas.openxmlformats.org/package/2006/relationships/metadata/core-properties";
-    public static final String EXTENDED_PROPERTIES_REL_TYPE = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/extended-properties";
-    public static final String CUSTOM_PROPERTIES_REL_TYPE = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/custom-properties";
-
     // OLE embeddings relation name
     public static final String OLE_OBJECT_REL_TYPE="http://schemas.openxmlformats.org/officeDocument/2006/relationships/oleObject";
 

src/ooxml/java/org/apache/poi/POIXMLDocumentPart.java

@@ -23,6 +23,7 @@ import java.net.URI;
 import org.apache.xmlbeans.XmlOptions;
 import org.apache.poi.util.POILogger;
 import org.apache.poi.util.POILogFactory;
+import org.apache.poi.openxml4j.exceptions.InvalidFormatException;
 import org.apache.poi.openxml4j.exceptions.OpenXML4JException;
 import org.apache.poi.openxml4j.opc.*;
 
@@ -86,6 +87,24 @@ public class POIXMLDocumentPart {
         this.packageRel = rel;
     }
     
+    /**
+     * When you open something like a theme, call this to
+     *  re-base the XML Document onto the core child of the
+     *  current core document 
+     */
+    protected final void rebase(OPCPackage pkg) throws InvalidFormatException {
+       PackageRelationshipCollection cores =
+          packagePart.getRelationshipsByType(PackageRelationshipTypes.CORE_DOCUMENT);
+       if(cores.size() != 1) {
+          throw new IllegalStateException(
+                "Tried to rebase using " + PackageRelationshipTypes.CORE_DOCUMENT +
+                " but found " + cores.size() + " parts of the right type"
+          );
+       }
+       packageRel = cores.getRelationship(0);
+       packagePart = POIXMLDocument.getTargetPart(pkg, packageRel);
+    }
+
     /**
      * Provides access to the underlying PackagePart
      *

src/ooxml/java/org/apache/poi/POIXMLProperties.java

@@ -28,6 +28,7 @@ import org.apache.poi.openxml4j.opc.OPCPackage;
 import org.apache.poi.openxml4j.opc.PackagePart;
 import org.apache.poi.openxml4j.opc.PackagePartName;
 import org.apache.poi.openxml4j.opc.PackageRelationshipCollection;
+import org.apache.poi.openxml4j.opc.PackageRelationshipTypes;
 import org.apache.poi.openxml4j.opc.PackagingURIHelper;
 import org.apache.poi.openxml4j.opc.TargetMode;
 import org.apache.poi.openxml4j.opc.internal.PackagePropertiesPart;
@@ -68,7 +69,7 @@ public class POIXMLProperties {
 
 		// Extended properties
 		PackageRelationshipCollection extRel =
-			pkg.getRelationshipsByType(POIXMLDocument.EXTENDED_PROPERTIES_REL_TYPE);
+			pkg.getRelationshipsByType(PackageRelationshipTypes.EXTENDED_PROPERTIES);
 		if(extRel.size() == 1) {
 			extPart = pkg.getPart( extRel.getRelationship(0));
 			org.openxmlformats.schemas.officeDocument.x2006.extendedProperties.PropertiesDocument props = org.openxmlformats.schemas.officeDocument.x2006.extendedProperties.PropertiesDocument.Factory.parse(
@@ -82,7 +83,7 @@ public class POIXMLProperties {
 
 		// Custom properties
 		PackageRelationshipCollection custRel =
-			pkg.getRelationshipsByType(POIXMLDocument.CUSTOM_PROPERTIES_REL_TYPE);
+			pkg.getRelationshipsByType(PackageRelationshipTypes.CUSTOM_PROPERTIES);
 		if(custRel.size() == 1) {
 			custPart = pkg.getPart( custRel.getRelationship(0));
 			org.openxmlformats.schemas.officeDocument.x2006.customProperties.PropertiesDocument props = org.openxmlformats.schemas.officeDocument.x2006.customProperties.PropertiesDocument.Factory.parse(

src/ooxml/java/org/apache/poi/openxml4j/opc/PackageRelationshipTypes.java

@@ -55,6 +55,11 @@ public interface PackageRelationshipTypes {
 	 */
 	String EXTENDED_PROPERTIES = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/extended-properties";
 
+	/**
+	 * Custom properties relationship type.
+	 */
+	String CUSTOM_PROPERTIES = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/custom-properties";
+	
 	/**
 	 * Core properties relationship type.
 	 */

src/ooxml/java/org/apache/poi/xslf/XSLFSlideShow.java

@@ -29,6 +29,7 @@ import org.apache.poi.openxml4j.opc.PackagePart;
 import org.apache.poi.openxml4j.opc.PackageRelationship;
 import org.apache.poi.openxml4j.opc.PackageRelationshipCollection;
 import org.apache.xmlbeans.XmlException;
+import org.openxmlformats.schemas.drawingml.x2006.main.ThemeDocument;
 import org.openxmlformats.schemas.presentationml.x2006.main.CTCommentList;
 import org.openxmlformats.schemas.presentationml.x2006.main.CTNotesSlide;
 import org.openxmlformats.schemas.presentationml.x2006.main.CTPresentation;
@@ -62,6 +63,7 @@ public class XSLFSlideShow extends POIXMLDocument {
    public static final String PRESENTATIONML_CONTENT_TYPE = "application/vnd.openxmlformats-officedocument.presentationml.slideshow.main+xml";
    public static final String PRESENTATIONML_TEMPLATE_CONTENT_TYPE = "application/vnd.openxmlformats-officedocument.presentationml.template.main+xml";
    public static final String PRESENTATION_MACRO_CONTENT_TYPE = "application/vnd.ms-powerpoint.presentation.macroEnabled.main+xml";
+   public static final String THEME_MANAGER_CONTENT_TYPE = "application/vnd.openxmlformats-officedocument.themeManager+xml";
 	public static final String NOTES_CONTENT_TYPE = "application/vnd.openxmlformats-officedocument.presentationml.notesSlide+xml";
 	public static final String SLIDE_CONTENT_TYPE = "application/vnd.openxmlformats-officedocument.presentationml.slide+xml";
 	public static final String SLIDE_LAYOUT_RELATION_TYPE = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/slideLayout";
@@ -77,6 +79,10 @@ public class XSLFSlideShow extends POIXMLDocument {
 	public XSLFSlideShow(OPCPackage container) throws OpenXML4JException, IOException, XmlException {
 		super(container);
 		
+		if(getCorePart().getContentType().equals(THEME_MANAGER_CONTENT_TYPE)) {
+		   rebase(getPackage());
+		}
+		
 		presentationDoc =
 			PresentationDocument.Factory.parse(getCorePart().getInputStream());
 		
@@ -112,6 +118,11 @@ public class XSLFSlideShow extends POIXMLDocument {
 	 */
     @Internal
 	public CTSlideIdList getSlideReferences() {
+       if(! getPresentation().isSetSldIdLst()) {
+          getPresentation().setSldIdLst(
+             CTSlideIdList.Factory.newInstance()   
+          );
+       }
        return getPresentation().getSldIdLst();
 	}
 	/**

src/ooxml/testcases/org/apache/poi/xslf/extractor/TestXSLFPowerPointExtractor.java

@@ -126,4 +126,53 @@ public class TestXSLFPowerPointExtractor extends TestCase {
         // Check comments are there
         assertTrue("Unable to find expected word in text\n" + text, text.contains("TEST"));
     }
+    
+    /**
+     * Test that we can get the text from macro enabled,
+     *  template, theme, slide enabled etc formats, as 
+     *  well as from the normal file
+     */
+    public void testDifferentSubformats() throws Exception {
+       POIDataSamples slTests = POIDataSamples.getSlideShowInstance();
+       String[] extensions = new String[] {
+             "pptx", "pptm", "ppsm", "ppsx",
+             "thmx", 
+             //"xps" // Doesn't have a core document
+       };
+       for(String extension : extensions) {
+          String filename = "testPPT." + extension;
+          xmlA = new XSLFSlideShow(OPCPackage.open(slTests.openResourceAsStream(filename)));
+          XSLFPowerPointExtractor extractor =
+             new XSLFPowerPointExtractor(xmlA);
+
+         String text = extractor.getText();
+         if(extension.equals("thmx")) {
+            // Theme file doesn't have any textual content
+            assertEquals(0, text.length());
+            continue;
+         }
+         
+         assertTrue(text.length() > 0);
+         assertTrue(
+               "Text missing for " + filename + "\n" + text, 
+               text.contains("Attachment Test")
+         );
+         assertTrue(
+               "Text missing for " + filename + "\n" + text, 
+               text.contains("This is a test file data with the same content")
+         );
+         assertTrue(
+               "Text missing for " + filename + "\n" + text, 
+               text.contains("content parsing")
+         );
+         assertTrue(
+               "Text missing for " + filename + "\n" + text, 
+               text.contains("Different words to test against")
+         );
+         assertTrue(
+               "Text missing for " + filename + "\n" + text, 
+               text.contains("Mystery")
+         );
+       }
+    }
 }