EPUB export + a bugfix

git-svn-id: svn://svn.code.sf.net/p/writer2latex/code/trunk@56 f0f2a975-2e09-46c8-9428-3b39399b9f3c

source/distro/Readme.txt

@@ -1,7 +1,7 @@
-Writer2LaTeX version 1.1.1 (development release)
+Writer2LaTeX version 1.1.2 (development release)
 ================================================
 
-This is the distribution of Writer2LaTeX version 1.1.1
+This is the distribution of Writer2LaTeX version 1.1.2
 
 Latest version can be found at the web site
   http://writer2latex.sourceforge.net
@@ -14,5 +14,5 @@ Bugs and feature requests should be reported to
   henrikjust (at) openoffice.org
   
 
-September 2009
+April 2010
 Henrik Just

source/distro/changelog.txt

@@ -2,6 +2,8 @@ Changelog for Writer2LaTeX version 1.0 -> 1.2
 
 ---------- version 1.1.2 ----------
 
+[w2x] Bugfix: Corrected problem with relative links (this affected package format only)
+
 [w2x] Added EPUB export
 
 [all] API change: The interface ConverterResult supports an additional method

source/java/writer2latex/api/ConverterFactory.java

@@ -20,7 +20,7 @@
  *
  *  All Rights Reserved.
  * 
- *  Version 1.2 (2010-03-29)
+ *  Version 1.2 (2010-03-31)
  *
  */
  
@@ -33,7 +33,7 @@ public class ConverterFactory {
 
     // Version information
     private static final String VERSION = "1.1.2";
-    private static final String DATE = "2010-03-29";
+    private static final String DATE = "2010-03-31";
 	
     /** Return the Writer2LaTeX version in the form
      *  (major version).(minor version).(patch level)<br/>

source/java/writer2latex/epub/ContainerWriter.java

@@ -38,11 +38,11 @@ import org.w3c.dom.Element;
 import writer2latex.xmerge.NewDOMDocument;
 
 /** This class creates the required META-INF/container.xml file for an EPUB package 
- *  (see http://www.idpf.org/ocf/ocf1.0/download/ocf10.htm)
+ *  (see http://www.idpf.org/ocf/ocf1.0/download/ocf10.htm).
  */
 public class ContainerWriter extends NewDOMDocument {
 	
-	public ContainerWriter(String sRootFile) {
+	public ContainerWriter() {
 		super("container", "xml");
 		
         // create DOM
@@ -67,7 +67,7 @@ public class ContainerWriter extends NewDOMDocument {
         container.appendChild(rootfiles);
         
         Element rootfile = contentDOM.createElement("rootfile");
-        rootfile.setAttribute("full-path", sRootFile);
+        rootfile.setAttribute("full-path", "OEBPS/book.opf");
         rootfile.setAttribute("media-type", "application/oebps-package+xml");
         rootfiles.appendChild(rootfile);
         

source/java/writer2latex/epub/EPUBConverter.java

@@ -44,6 +44,7 @@ public final class EPUBConverter extends Xhtml11Converter {
     }
 	
     @Override public ConverterResult convert(InputStream is, String sTargetFileName) throws IOException {
+    	setOPS(true);
     	ConverterResult xhtmlResult = super.convert(is, sTargetFileName);
     	
     	ConverterResultImpl epubResult = new ConverterResultImpl();

source/java/writer2latex/epub/EPUBWriter.java

@@ -20,7 +20,7 @@
  *
  *  All Rights Reserved.
  * 
- *  version 1.2 (2010-03-29)
+ *  version 1.2 (2010-03-31)
  *
  */
 
@@ -30,6 +30,7 @@ import java.io.ByteArrayOutputStream;
 import java.io.IOException;
 import java.io.OutputStream;
 import java.util.Iterator;
+import java.util.UUID;
 import java.util.zip.ZipEntry;
 import java.util.zip.ZipOutputStream;
 
@@ -37,8 +38,9 @@ import writer2latex.api.ConverterResult;
 import writer2latex.api.OutputFile;
 import writer2latex.util.Misc;
 
-/** This class repackages an XHTML document into EPUB format
+/** This class repackages an XHTML document into EPUB format.
- * 
+ *  Some filenames are hard wired in this implementation: The main directory is OEBPS and
+ *  the OPF and NCX files are book.opf and book.ncx respectively 
  */
 public class EPUBWriter implements OutputFile {
 	
@@ -66,9 +68,12 @@ public class EPUBWriter implements OutputFile {
 	}
 
 	public void write(OutputStream os) throws IOException {
+		// Create a universal unique ID
+		String sUUID = UUID.randomUUID().toString(); 
+		
 		ZipOutputStream zos = new ZipOutputStream(os);
 		
-		// Write MIME type as first entry
+		// Write uncompressed MIME type as first entry
 		ZipEntry mimeEntry = new ZipEntry("mimetype");
 		mimeEntry.setMethod(ZipEntry.STORED);
 		mimeEntry.setCrc(0x2CAB616F);
@@ -78,27 +83,27 @@ public class EPUBWriter implements OutputFile {
 		zos.closeEntry();
 		
 		// Write container entry next
-		OutputFile containerWriter = new ContainerWriter("OEBPS/book.opf");
+		OutputFile containerWriter = new ContainerWriter();
 		ZipEntry containerEntry = new ZipEntry("META-INF/container.xml");
 		zos.putNextEntry(containerEntry);
 		writeZipEntry(containerWriter,zos);
 		zos.closeEntry();
 		
 		// Then manifest
-		OutputFile manifest = new OPFWriter(xhtmlResult, "xxx", "book.ncx", "book.opf");
+		OutputFile manifest = new OPFWriter(xhtmlResult, sUUID);
 		ZipEntry manifestEntry = new ZipEntry("OEBPS/book.opf");
 		zos.putNextEntry(manifestEntry);
 		writeZipEntry(manifest,zos);
 		zos.closeEntry();
 		
 		// And content table
-		OutputFile ncx = new NCXWriter(xhtmlResult, "xxx", "book.ncx");
+		OutputFile ncx = new NCXWriter(xhtmlResult, sUUID);
 		ZipEntry ncxEntry = new ZipEntry("OEBPS/book.ncx");
 		zos.putNextEntry(ncxEntry);
 		writeZipEntry(ncx,zos);
 		zos.closeEntry();
 		
-		// Finally XHTML content in the OEBPS sub directory
+		// Finally XHTML content
 		Iterator<OutputFile> iter = xhtmlResult.iterator();
 		while (iter.hasNext()) {
 			OutputFile file = iter.next();

source/java/writer2latex/epub/NCXWriter.java

@@ -20,7 +20,7 @@
  *
  *  All Rights Reserved.
  * 
- *  version 1.2 (2010-03-29)
+ *  version 1.2 (2010-03-30)
  *
  */
 
@@ -39,17 +39,15 @@ import org.w3c.dom.Element;
 
 import writer2latex.api.ContentEntry;
 import writer2latex.api.ConverterResult;
-import writer2latex.util.Misc;
 import writer2latex.xmerge.NewDOMDocument;
 
 /** This class creates the required NXC file for an EPUB document
- *  (see http://www.idpf.org/2007/opf/OPF_2.0_final_spec.html#Section2.4)
+ *  (see http://www.idpf.org/2007/opf/OPF_2.0_final_spec.html#Section2.4).
- * 
  */
 public class NCXWriter extends NewDOMDocument {
 	
-	public NCXWriter(ConverterResult cr, String sUUID, String sFileName) {
+	public NCXWriter(ConverterResult cr, String sUUID) {
-		super(Misc.removeExtension(sFileName), "ncx");
+		super("book", "ncx");
 		
         // create DOM
         Document contentDOM = null;
@@ -64,8 +62,6 @@ public class NCXWriter extends NewDOMDocument {
             throw new RuntimeException(t);
         }
         
-        System.out.println("populating the ncx");
-        
         // Populate the DOM tree
         Element ncx = contentDOM.getDocumentElement();
         ncx.setAttribute("version", "2005-1");
@@ -83,7 +79,7 @@ public class NCXWriter extends NewDOMDocument {
         
         Element depth = contentDOM.createElement("meta");
         depth.setAttribute("name","dtb:depth");
-        depth.setAttribute("content", "1");
+        // Setting the content attribute later
         head.appendChild(depth);
         
         Element totalPageCount = contentDOM.createElement("meta");
@@ -107,35 +103,57 @@ public class NCXWriter extends NewDOMDocument {
         Element navMap = contentDOM.createElement("navMap");
         ncx.appendChild(navMap);
         
+        Element currentContainer = ncx;
+        int nCurrentLevel = 0;
+        int nDepth = 0;
         int nPlayOrder = 0;
+        
         Iterator<ContentEntry> content = cr.getContent().iterator();
         while (content.hasNext()) {
         	ContentEntry entry = content.next();
-        	//if (entry.getLevel()==1) {
+        	int nEntryLevel = Math.max(entry.getLevel(), 1);
-        	System.out.println("Found content entry "+entry.getTitle());
+        	
-        		Element navPoint = contentDOM.createElement("navPoint");
+        	if (nEntryLevel<nCurrentLevel) {
-        		navPoint.setAttribute("playOrder", Integer.toString(++nPlayOrder));
+        		// Return to higher level
-        		navPoint.setAttribute("id", "text"+nPlayOrder);
+        		for (int i=nEntryLevel; i<nCurrentLevel; i++) {
-        		navMap.appendChild(navPoint);
+        			currentContainer = (Element) currentContainer.getParentNode();
-        		
+        		}
-        		Element navLabel = contentDOM.createElement("navLabel");
+        		nCurrentLevel = nEntryLevel;
-        		navPoint.appendChild(navLabel);
+        	}
-        		Element navLabelText = contentDOM.createElement("text");
+        	else if (nEntryLevel>nCurrentLevel) {
-        		navLabel.appendChild(navLabelText);
+        		// To lower level (always one step; a jump from e.g. heading 1 to heading 3 in the document
-        		navLabelText.appendChild(contentDOM.createTextNode(entry.getTitle()));
+        		// is considered an error)
-        		
+        		currentContainer = (Element) currentContainer.getLastChild();
-        		Element navPointContent = contentDOM.createElement("content");
+        		nCurrentLevel++;
-    			String sHref = entry.getFile().getFileName();
+        	}
-    			if (entry.getTarget()!=null) { sHref+="#"+entry.getTarget(); }
+        	
-        		navPointContent.setAttribute("src", sHref);
+        	Element navPoint = contentDOM.createElement("navPoint");
-        		navPoint.appendChild(navPointContent);
+        	navPoint.setAttribute("playOrder", Integer.toString(++nPlayOrder));
-        	//}
+        	navPoint.setAttribute("id", "text"+nPlayOrder);
+        	currentContainer.appendChild(navPoint);
+
+        	Element navLabel = contentDOM.createElement("navLabel");
+        	navPoint.appendChild(navLabel);
+        	Element navLabelText = contentDOM.createElement("text");
+        	navLabel.appendChild(navLabelText);
+        	navLabelText.appendChild(contentDOM.createTextNode(entry.getTitle()));
+
+        	Element navPointContent = contentDOM.createElement("content");
+        	String sHref = entry.getFile().getFileName();
+        	if (entry.getTarget()!=null) { sHref+="#"+entry.getTarget(); }
+        	navPointContent.setAttribute("src", sHref);
+        	navPoint.appendChild(navPointContent);
+        	
+        	nDepth = Math.max(nDepth, nCurrentLevel);
         }
         
+        if (nDepth==0) {
+        	// TODO: We're in trouble: The document has no headings
+        }
+        
+        depth.setAttribute("content", Integer.toString(nDepth));
+        
         setContentDOM(contentDOM);
         
-        System.out.println("finished populating the ncx");
-
-        
 	}
 }

source/java/writer2latex/epub/OPFWriter.java

@@ -41,16 +41,14 @@ import org.w3c.dom.Element;
 import writer2latex.api.ContentEntry;
 import writer2latex.api.ConverterResult;
 import writer2latex.api.OutputFile;
-import writer2latex.util.Misc;
 import writer2latex.xmerge.NewDOMDocument;
 
-/** This class writes an OPF-file for an EPUB document (see http://www.idpf.org/2007/opf/OPF_2.0_final_spec.html)
+/** This class writes an OPF-file for an EPUB document (see http://www.idpf.org/2007/opf/OPF_2.0_final_spec.html).
- * 
  */
 public class OPFWriter extends NewDOMDocument {
 
-	public OPFWriter(ConverterResult cr, String sUUID, String sNcxFileName, String sFileName) {
+	public OPFWriter(ConverterResult cr, String sUUID) {
-		super(Misc.removeExtension(sFileName), "opf");
+		super("book", "opf");
 		
         // create DOM
         Document contentDOM = null;
@@ -123,7 +121,7 @@ public class OPFWriter extends NewDOMDocument {
         }
         
         Element item = contentDOM.createElement("item");
-        item.setAttribute("href", sNcxFileName);
+        item.setAttribute("href", "book.ncx");
         item.setAttribute("media-type", "application/x-dtbncx+xml");
         item.setAttribute("id", "ncx");
         manifest.appendChild(item);

source/java/writer2latex/latex/DrawConverter.java

@@ -20,7 +20,7 @@
  *
  *  All Rights Reserved.
  * 
- *  Version 1.2 (2010-03-15)
+ *  Version 1.2 (2010-03-29)
  *
  */
  
@@ -343,11 +343,7 @@ public class DrawConverter extends ConverterHelper {
         if (node.hasAttribute(XMLString.XLINK_HREF) && !ofr.isInPackage(sHref)) {
             // Linked image is not yet handled by ImageLoader. This is a temp.
             // solution (will go away when ImageLoader is finished)
-            sFileName = sHref;
+            sFileName = ofr.fixRelativeLink(sHref);
-            // In OpenDocument package format ../ means "leave the package"
-            if (ofr.isOpenDocument() && ofr.isPackageFormat() && sFileName.startsWith("../")) {
-                sFileName=sFileName.substring(3);
-            }
             int nExtStart = sHref.lastIndexOf(".");
             String sExt = nExtStart>=0 ? sHref.substring(nExtStart).toLowerCase() : "";
             // Accept only relative filenames and supported filetypes:

source/java/writer2latex/office/OfficeReader.java

@@ -16,11 +16,11 @@
  *  Foundation, Inc., 59 Temple Place, Suite 330, Boston,
  *  MA  02111-1307  USA
  *
- *  Copyright: 2002-2008 by Henrik Just
+ *  Copyright: 2002-2010 by Henrik Just
  *
  *  All Rights Reserved.
  * 
- *  Version 1.2 (2008-09-30)
+ *  Version 1.2 (2010-03-29)
  *
  */
 
@@ -334,6 +334,19 @@ public class OfficeReader {
         if (sUrl.startsWith("./")) { sUrl=sUrl.substring(2); }
         return oooDoc.getEmbeddedObject(sUrl)!=null; 
     } 
+    
+    /** In OpenDocument package format ../ means "leave the package".
+     *  Consequently this prefix must be removed to obtain a valid link 
+     *  
+     * @param sLink
+     * @return the corrected link
+     */
+    public String fixRelativeLink(String sLink) {
+        if (isOpenDocument() && isPackageFormat() && sLink.startsWith("../")) {
+            return sLink.substring(3);
+        }
+        return sLink;
+    }
 	
     ///////////////////////////////////////////////////////////////////////////
     // Accessor methods

source/java/writer2latex/xhtml/Converter.java

@@ -20,7 +20,7 @@
  *
  *  All Rights Reserved.
  * 
- *  Version 1.2 (2010-03-25)
+ *  Version 1.2 (2010-03-31)
  *
  */
 
@@ -44,7 +44,6 @@ import org.w3c.dom.NodeList;
 import org.w3c.dom.Text;
 
 import writer2latex.api.Config;
-import writer2latex.api.ContentEntry;
 import writer2latex.api.ConverterFactory;
 //import writer2latex.api.ConverterResult;
 import writer2latex.base.ContentEntryImpl;
@@ -83,6 +82,7 @@ public class Converter extends ConverterBase {
 
     // The xhtml output file(s)
     protected int nType = XhtmlDocument.XHTML10; // the doctype
+    private boolean bOPS = false; // Do we need to be OPS conforming?
     Vector<XhtmlDocument> outFiles;
     private int nOutFileIndex;
     private XhtmlDocument htmlDoc; // current outfile
@@ -157,15 +157,18 @@ public class Converter extends ConverterBase {
     protected Node importNode(Node node, boolean bDeep) { return htmlDOM.importNode(node,bDeep); }
 	
     protected L10n getL10n() { return l10n; }
+    
+    public void setOPS(boolean b) { bOPS = true; }
+    
+    public boolean isOPS() { return bOPS; }
 	
-    // override
+    @Override public void convertInner() throws IOException {      
-    public void convertInner() throws IOException {      
         sTargetFileName = Misc.trimDocumentName(sTargetFileName,XhtmlDocument.getExtension(nType));
 		
         outFiles = new Vector<XhtmlDocument>();
         nOutFileIndex = -1;
 
-        bNeedHeaderFooter = ofr.isSpreadsheet() || ofr.isPresentation() || config.getXhtmlSplitLevel()>0 || config.getXhtmlUplink().length()>0;
+        bNeedHeaderFooter = !bOPS && (ofr.isSpreadsheet() || ofr.isPresentation() || config.getXhtmlSplitLevel()>0 || config.getXhtmlUplink().length()>0);
 
         l10n = new L10n();
         
@@ -612,6 +615,8 @@ public class Converter extends ConverterBase {
         }
         else { // external link
             anchor = htmlDOM.createElement("a");
+            
+            sHref = ofr.fixRelativeLink(sHref);
 	
             // Workaround for an OOo problem:
             if (sHref.indexOf("?")==-1) { // No question mark

source/java/writer2latex/xhtml/TextConverter.java

@@ -20,7 +20,7 @@
  *
  *  All Rights Reserved.
  * 
- *  Version 1.2 (2010-03-26)
+ *  Version 1.2 (2010-03-29)
  *
  */
 
@@ -571,11 +571,14 @@ public class TextConverter extends ConverterHelper {
             			
             // Add to toc
             if (!bInToc) {
-                converter.addTarget(heading,"toc"+(++nTocIndex));
+            	String sTarget = "toc"+(++nTocIndex);
+                converter.addTarget(heading,sTarget);
                 
-                // Add in external content
+                // Add in external content. For single file output we include all level 1 headings + their target
-                if (nLevel<=nSplit) {
+                // For multi-file output, the included heading levels depends on the split leve, and we don't include targets 
-                	converter.addContentEntry(sLabel+(sLabel.length()>0 ? " " : "")+Misc.getPCDATA(onode), nLevel, null);
+                if (nLevel<=Math.max(nSplit,1)) {
+                	converter.addContentEntry(sLabel+(sLabel.length()>0 ? " " : "")+Misc.getPCDATA(onode), nLevel,
+                			nSplit==0 ? sTarget : null);
                 }
 
                 // Add to real toc

source/readme-source.txt

@@ -1,7 +1,7 @@
-Writer2LaTeX source version 1.1.1
+Writer2LaTeX source version 1.1.2
 =================================
 
-Writer2LaTeX is (c) 2002-2009 by Henrik Just.
+Writer2LaTeX is (c) 2002-2010 by Henrik Just.
 The source is available under the terms and conditions of the
 GNU LESSER GENERAL PUBLIC LICENSE, version 2.1.
 Please see the file COPYING.TXT for details.
@@ -13,7 +13,7 @@ Overview
 The source of Writer2LaTeX consists of three major parts:
 
 * A general purpose java library for converting OpenDocument files into LaTeX,
-  BibTeX, xhtml and xhtml+MathML
+  BibTeX, xhtml, xhtml+MathML and EPUB
   This is to be found in the packages writer2latex.* and should only be used
   through the provided api writer2latex.api.*
 * A command line utility writer2latex.Application
@@ -44,15 +44,15 @@ Writer2LaTeX uses Ant version 1.5 or later (http://ant.apache.org) to build.
 Some java libraries from OOo are needed to build the filter part of Writer2LaTeX,
 these are jurt.jar, unoil.jar, ridl.jar and juh.jar.
 
-To make these files available for the compiler, edit the file build.xml in
+To make these files available for the compiler, edit the file build.xml
-the writer2latex09 directory as follows:
+as follows:
 
 The lines
-	<property name="OFFICE_HOME" location="/opt/openoffice.org/basis3.0" />
+	<property name="OFFICE_CLASSES" location="/usr/share/java/openoffice" />
-	<property name="URE_HOME" location="/opt/openoffice.org/ure" />
+	<property name="URE_CLASSES" location="/usr/share/java/openoffice" />
-should be modified to point to your OOo installation
+should be modified to the directories where your OOo installation keeps these files
 
-To build, open a command shell, navigate to the writer2latex09 directory and type
+To build, open a command shell, navigate to the source directory and type
 
 ant oxt
 
@@ -72,7 +72,7 @@ In addition to oxt, the build file supports the following targets:
     clean
 
 
-Henrik Just, November 2009
+Henrik Just, April 2010
 
 
 Thanks to Michael Niedermair for writing the original ant build file