When importing a word document and splitting by headers, the page tree that gets created is alphabetized. What we are expecting to happen is that the page tree would be in the same order in which the headers appear in the document.
To reproduce, simply import a word document which can be split by multiple headers that do not appear in alphabetical order within the document. Once the document is imported, notice that the page tree is alphabetical in order as opposed to their order within the document.
I have attached an example Word document to test with.
- Import Brian.docx to Confluence and use Split by heading functionality.
- Observe pages tree generated in sidebar.
- Page tree will be ordered properly (as in document).
The issue is not reproducible with 5.10.8.
It looks like problem is introduced in Confluence version 6.0.6 and plugin OfficeConnector-5.0.3.jar. The issue is not present in plugin 5.0.2 and Confluence 6.0.5.
There are 2 workaround at this point:
- Manually sort and reorder pages.
- Downgrade the plugin:
- Stop confluence
- Find the folder <confluence-install>/confluence/WEB-INF/atlassian-bundled-plugins/ on production
- Backup and move OfficeConnector-5.x.x.jar from this folder
- Download Confluence 5.10.8 zip.
- Extract from the same folder OfficeConnector-5.0.2.jar. and copy it to the same folder on production.
- Remove the following folders:
- Start Confluence
Note: The work around (2) has been confirmed to work up up to Confluence 6.5.2 but does not work for Confluence 6.7.1 (Confluence 6.7.1 will crash when importing Word Documents when using the older OfficeConnector-5.0.2.jar)