Image export improvements and bugfixes

git-svn-id: svn://svn.code.sf.net/p/writer2latex/code/trunk@177 f0f2a975-2e09-46c8-9428-3b39399b9f3c

source/distro/changelog.txt

@@ -2,6 +2,13 @@ Changelog for Writer2LaTeX version 1.2 -> 1.4
 
 ---------- version 1.4 beta ----------
 
+[w2x] Bugfix: Avoid null pointer exception in HTML5 export if embed_svg is true and an image is used more than once
+
+[all] An image that is used more than once in the source document is now only exported once. This was already the
+  case for the package format (since 1.3.2), but now also for flat XML.
+
+[w2x] The The option inline_svg has been renamed (again) to embed_svg, and the default is changed to false
+
 [w2x] The experimental option zen_hack has been removed
 
 [w2x] For templates it is no longer required that the footer, header and panel are contained in a div element.

source/java/org/openoffice/da/comp/writer2xhtml/ConfigurationDialog.java

@@ -20,7 +20,7 @@
 *
 *  All Rights Reserved.
 * 
-*  Version 1.4 (2014-09-19)
+*  Version 1.4 (2014-09-23)
 *
 */ 
 
@@ -495,14 +495,14 @@ public class ConfigurationDialog extends ConfigurationDialogBase implements XSer
     		listBoxFromConfig(dlg, "Formulas", "formulas", sFormulaValues, (short) 0);
     		textFieldFromConfig(dlg, "EndnotesHeading", "endnotes_heading");
     		textFieldFromConfig(dlg, "FootnotesHeading", "footnotes_heading");
-    		checkBoxFromConfig(dlg, "InlineSvg", "inline_svg");
+    		checkBoxFromConfig(dlg, "EmbedSvg", "embed_svg");
     	}
     	
     	@Override protected void getControls(DialogAccess dlg) {
     		listBoxToConfig(dlg, "Formulas", "formulas", sFormulaValues);
     		textFieldToConfig(dlg, "EndnotesHeading", "endnotes_heading");
     		textFieldToConfig(dlg, "FootnotesHeading", "footnotes_heading");
-    		checkBoxToConfig(dlg, "InlineSvg", "inline_svg");
+    		checkBoxToConfig(dlg, "EmbedSvg", "embed_svg");
     	}
     	
     	@Override protected boolean handleEvent(DialogAccess dlg, String sMethod) {

source/java/writer2latex/base/ImageConverter.java

@@ -20,17 +20,22 @@
  *
  *  All Rights Reserved.
  * 
- *  Version 1.4 (2014-09-16)
+ *  Version 1.4 (2014-09-24)
  *
  */
 
 package writer2latex.base;
 
+import java.security.MessageDigest;
+import java.security.NoSuchAlgorithmException;
 import java.text.DecimalFormat;
 import java.text.NumberFormat;
+import java.util.Arrays;
 import java.util.HashMap;
 import java.util.HashSet;
 
+import javax.xml.bind.DatatypeConverter;
+
 import org.w3c.dom.Element;
 import org.w3c.dom.Node;
 import org.w3c.dom.NodeList;
@@ -181,12 +186,6 @@ public final class ImageConverter {
     		 // The file name was not used
     		nImageCount--;
     	}
-    	else if (node.hasAttribute(XMLString.XLINK_HREF)) {
-    		// This is an embedded image we meet for the first time.
-    		// Recycle it on behalf of the original image node.
-    		String sHref = node.getAttribute(XMLString.XLINK_HREF);
-    		recycledImages.put(sHref, new BinaryGraphicsDocument(bgd));
-    	}
     	return bgd;
     }
     
@@ -197,6 +196,7 @@ public final class ImageConverter {
         String sExt = null;
     	String sMIME = null;
     	byte[] blob = null;
+    	String sId = null;
     	
     	// First try to extract the image using the xlink:href attribute
     	if (node.hasAttribute(XMLString.XLINK_HREF)) {
@@ -224,6 +224,8 @@ public final class ImageConverter {
 	            	if (bDestructive) {
 	            		object.dispose();
 	            	}
+	            	// We got an image, define ID for recycling
+	            	sId = sHref;
 	            }
 	            else {
 	                // This is a linked image
@@ -249,11 +251,18 @@ public final class ImageConverter {
 	                }
 	            }
 	            blob = Base64.decode(buf.toString());
+    			// We may have seen this image before, return the recycled version
+	            String sId1 = createId(blob);
+    			if (recycledImages.containsKey(sId1)) {
+    				return recycledImages.get(sId1);
+    			}
     			sMIME = MIMETypes.getMagicMIMEType(blob);
 	            sExt = MIMETypes.getFileExtension(sMIME);
 	            if (bDestructive) {
 	            	node.removeChild(obd);
 	            }
+	            // We got an image, define ID for recycling
+	            sId = sId1;
 	        }
 	        else {
 	        	// There is no image data
@@ -303,11 +312,13 @@ public final class ImageConverter {
         }
         
         // Create the result
-
         if (isAcceptedFormat(sMIME) || bAcceptOtherFormats) {
         	String sFileName = sName+sExt;
             BinaryGraphicsDocument bgd = new BinaryGraphicsDocument(sFileName,sMIME);
             bgd.setData(blob,isAcceptedFormat(sMIME));
+            if (sId!=null) {
+        		recycledImages.put(sId, new BinaryGraphicsDocument(bgd));
+            }
             return bgd;
         }
         else {
@@ -323,4 +334,17 @@ public final class ImageConverter {
     	return null;
     }
     
+    // Create a fingerprint of a blob. The fingerprint concatenates the MD5 hash with the first 10 bytes of the blob.
+    private String createId(byte[] blob) {
+    	MessageDigest md;
+    	try {
+			md = MessageDigest.getInstance("MD5");
+		} catch (NoSuchAlgorithmException e) {
+			// This would be surprising
+			return null;
+		}
+    	return DatatypeConverter.printHexBinary(md.digest(blob))
+    			+DatatypeConverter.printHexBinary(Arrays.copyOf(blob, 10));
+    }
+    
 }

source/java/writer2latex/xhtml/DrawConverter.java

@@ -20,7 +20,7 @@
  *
  *  All Rights Reserved.
  * 
- *  Version 1.4 (2014-09-05)
+ *  Version 1.4 (2014-09-24)
  *
  */
  
@@ -38,9 +38,9 @@
   * export notes (part of draw-page)
   * export list-style-image for image bullets!
   */
-
 package writer2latex.xhtml;
 
+import java.util.HashMap;
 import java.util.Iterator;
 import java.util.Vector;
 
@@ -54,8 +54,6 @@ import org.w3c.dom.Node;
 import org.w3c.dom.NodeList;
 import org.w3c.dom.Element;
 
-//import writer2latex.xmerge.EmbeddedBinaryObject;
-
 import writer2latex.util.Misc;
 import writer2latex.util.CSVList;
 import writer2latex.util.SimpleXMLParser;
@@ -101,7 +99,10 @@ public class DrawConverter extends ConverterHelper {
     private int nImageSize;
     private String sImageSplit;
     private boolean bCoverImage;
-    private boolean bUseSVG;
+    private boolean bEmbedSVG;
+
+    // Embedded SVG images for recycling are collected here
+    private HashMap<String,Element> svgImages = new HashMap<String,Element>();
     
     // Frames in spreadsheet documents are collected here
     private Vector<Element> frames = new Vector<Element>();
@@ -128,7 +129,7 @@ public class DrawConverter extends ConverterHelper {
         nImageSize = config.imageSize();
         sImageSplit = config.imageSplit();
         bCoverImage = config.coverImage();
-        bUseSVG = config.inlineSVG();
+        bEmbedSVG = config.embedSVG();
     }
     
     ///////////////////////////////////////////////////////////////////////
@@ -486,14 +487,25 @@ public class DrawConverter extends ConverterHelper {
         
         // Create the image (sFileName contains the file name)
         Element imageElement = null;
-        if (converter.nType==XhtmlDocument.HTML5 && bUseSVG && bgd!=null && MIMETypes.SVG.equals(bgd.getMIMEType())) {
+        if (converter.nType==XhtmlDocument.HTML5 && bEmbedSVG && bgd!=null && MIMETypes.SVG.equals(bgd.getMIMEType())) {
         	// In HTML5 we may embed SVG images directly in the document
+        	if (bgd.isRecycled()) {
+        		// Get from the cache (actually we are certain it is there)
+        		if (svgImages.containsKey(bgd.getFileName())) {
+					Element elm = hnodeInline!=null ? hnodeInline : hnodeBlock;
+					imageElement = (Element) elm.getOwnerDocument().importNode(svgImages.get(bgd.getFileName()), true);
+        		}
+        	}
+        	else {
+        		// Parse the data
 	        	byte[] blob = bgd.getData();
 	        	try {
 					Document dom = SimpleXMLParser.parse(new ByteArrayInputStream(blob));
 					if (dom!=null) {
 						Element elm = hnodeInline!=null ? hnodeInline : hnodeBlock;
 						imageElement = (Element) elm.getOwnerDocument().importNode(dom.getDocumentElement(), true);
+						// Store in cache in case we need to recycle
+						svgImages.put(bgd.getFileName(), imageElement);
 					}
 					else {
 						System.out.println("Failed to parse SVG");
@@ -507,6 +519,7 @@ public class DrawConverter extends ConverterHelper {
 					System.out.println("SAXException parsing SVG");
 				}
         	}
+        }
         else {
         	// In all other cases, create an img element
             if (bgd!=null && !bgd.isLinked() && !bgd.isRecycled()) { converter.addDocument(bgd); }

source/java/writer2latex/xhtml/XhtmlConfig.java

@@ -143,7 +143,7 @@ public class XhtmlConfig extends writer2latex.base.ConfigBase {
     private static final int SPLIT_AFTER = 40;
     private static final int IMAGE_SPLIT = 41;
     private static final int COVER_IMAGE = 42;
-    private static final int INLINE_SVG = 43;
+    private static final int EMBED_SVG = 43;
     private static final int USE_MATHJAX = 44;
     private static final int CALC_SPLIT = 45;
     private static final int DISPLAY_HIDDEN_SHEETS = 46;
@@ -273,7 +273,7 @@ public class XhtmlConfig extends writer2latex.base.ConfigBase {
         };
         options[IMAGE_SPLIT] = new Option("image_split","none");
         options[COVER_IMAGE] = new BooleanOption("cover_image","false");
-        options[INLINE_SVG] = new BooleanOption("inline_svg","true");
+        options[EMBED_SVG] = new BooleanOption("embed_svg","false");
         options[USE_MATHJAX] = new BooleanOption("use_mathjax","false");
         options[CALC_SPLIT] = new BooleanOption("calc_split","false");
         options[DISPLAY_HIDDEN_SHEETS] = new BooleanOption("display_hidden_sheets", "false");
@@ -399,7 +399,7 @@ public class XhtmlConfig extends writer2latex.base.ConfigBase {
     public int splitAfter() { return ((IntegerOption) options[SPLIT_AFTER]).getValue(); }
     public String imageSplit() { return options[IMAGE_SPLIT].getString(); }
     public boolean coverImage() { return ((BooleanOption) options[COVER_IMAGE]).getValue(); }
-    public boolean inlineSVG() { return ((BooleanOption) options[INLINE_SVG]).getValue(); }
+    public boolean embedSVG() { return ((BooleanOption) options[EMBED_SVG]).getValue(); }
     public boolean useMathJax() { return ((BooleanOption) options[USE_MATHJAX]).getValue(); }
     public boolean xhtmlCalcSplit() { return ((BooleanOption) options[CALC_SPLIT]).getValue(); }
     public boolean xhtmlDisplayHiddenSheets() { return ((BooleanOption) options[DISPLAY_HIDDEN_SHEETS]).getValue(); }

source/oxt/writer2xhtml/W2XDialogs2/Content.xdl

@@ -18,6 +18,6 @@
   <dlg:text dlg:id="FootnotesHeadingLabel" dlg:tab-index="6" dlg:left="10" dlg:top="64" dlg:width="110" dlg:height="12" dlg:value="Footnotes heading"/>
   <dlg:textfield dlg:id="FootnotesHeading" dlg:tab-index="7" dlg:left="130" dlg:top="62" dlg:width="120" dlg:height="12" dlg:help-url="org.openoffice.da.writer2xhtml.oxt:FootnotesHeading"/>
   <dlg:text dlg:id="SvgLabel" dlg:tab-index="8" dlg:left="5" dlg:top="80" dlg:width="245" dlg:height="12" dlg:value="SVG images (HTML5 only)"/>
-  <dlg:checkbox dlg:id="InlineSvg" dlg:tab-index="9" dlg:left="10" dlg:top="92" dlg:width="240" dlg:height="12" dlg:value="Embed SVG images in the HTML document" dlg:checked="false" dlg:help-url="org.openoffice.da.writer2xhtml.oxt:InlineSvg"/>
+  <dlg:checkbox dlg:id="EmbedSvg" dlg:tab-index="9" dlg:left="10" dlg:top="92" dlg:width="240" dlg:height="12" dlg:value="Embed SVG images in the HTML document" dlg:checked="false" dlg:help-url="org.openoffice.da.writer2xhtml.oxt:EmbedSvg"/>
  </dlg:bulletinboard>
 </dlg:window>

source/oxt/writer2xhtml/help/en/org.openoffice.da.writer2xhtml.oxt/Configuration/Content.xhp

@@ -54,7 +54,7 @@
 
         <paragraph role="heading" level="2" xml-lang="en-US">SVG images</paragraph>
 
-        <bookmark xml-lang="en-US" branch="hid/org.openoffice.da.writer2xhtml.oxt:InlineSvg" id="bm_configinlinesvg"/>
+        <bookmark xml-lang="en-US" branch="hid/org.openoffice.da.writer2xhtml.oxt:EmbedSvg" id="bm_configembedsvg"/>
 
         <paragraph role="heading" level="3" xml-lang="en-US">Embed SVG images in the HTML document</paragraph>
         <paragraph role="paragraph" xml-lang="en-US">If you check this option, SVG images will be embedded directly