Hello odf users, I would like to ask you to test the latest sources with your environment as there had been several updates.
While working on the patch for https://issues.apache.org/jira/browse/ODFTOOLKIT-413 over the holiday season, I have started to add several other fixes to it, I like to mention: 1. The main fix of https://issues.apache.org/jira/browse/ODFTOOLKIT-413 The representation of the ODF manifest.xml was held twice at run-time. Once as DOM tree and the children of the root element (file entries) as well within a hash map for fast access. Those two instances were once not in synch. Fixed this. 2. There was one fix missing for applying the same solution of https://issues.apache.org/jira/browse/ODFTOOLKIT-396. Fix now applied at ..pkg.OdfElement method public Node cloneNode(boolean deep) another differentiation of attribute with and without namespace was missing. 3. When building under Ubuntu, I had to update the multi-schema-validator implementation in the pom.xml as the prior was no longer able to be downloaded. 4. As I often work offline, I have enabled the offline build without Internet by adding the missing ODF schema 5. To be able to run/debug only one test from the context menu in Netbeans, I had to update the version of the maven-wagon in the pom.xml of various projects. 6. In addition many fixes for the OdfPackage class based on my prior work on the fork of Open-XChange were added. Sorry, it was too time consuming to split those in a clean way into separate patches, while working at the solution and stumbling over and over again of a small problem with an easy solution: 1. Splitting Exception into IOException and SAXException 2. Applying JDK 5 Generics several times 3. Removing potential vulnerability: see https://www.owasp.org/index.php/XML_External_Entity_%28XXE%29_Processing 4. Several minor fixes.. 7. Finally, I am sorry there is quite some line changing noise due to an accidentally used format function of my Netbeans IDE, which seems to use a different format algorithm to the one being used in Eclipse. Does anybody have a solution to this format problem? Shall we use a Maven plugin for this automatically normalizing the source code during build time? The build of the patch was tested under Ubuntu and Windows10. Best regards, Svante