Hello,
After many tests on the main features expected for a migration from Confluence to XWiki, I am now looking closely into the content of the pages imported with each space.
There are about 50 spaces to be imported, with several years of work in it, so I am still testing, and looking at the results compared to expected results.
I find top level pages contents from spaces imported from Confluence to XWiki missing, (all tests from 14.10 up to 15.10 LTS and 10.6.0), and I noticed also blog pages contents missing. (I am still in the process of checking blog pages in other spaces).
The context :
Debian installs in Virtual Machines, now we use Xjetty, and before, we were using Tomcat9.
The tool used to perform importation is the Filter Streams Converter, manual methode. (Confluence XML Application extension).
The Filter Streams Converter tool has come with different changes in the options along the XWiki versions, and I have done my best to follow up, and got recently useful answers to whats and hows options I should use and activate.
One of these options is a choice concerning the names of the spaces : Should they be the ID of the space (which means not changing the name of the space by default in the options), or should they be the Name of the space (which then means write true in the relevant field).
I have set it up to true only recently in my last tests. I have imported a bunch of spaces currently used in the v16.6.0 XWiki.
Wether I used true or false, import it with ID of the spaces for titles or import it with Space name for title, the result regarding missing contents in top level pages is the same, the content is missing, and also in Blogs the contents are missing.
I suspect the LINK MAPPING could perhaps help, as there might also be some correct links to pages missing, but I don’t yet know how to choose a format amongst the ones proposed.
I noticed some top level pages imported in XWiki present themselves with a warning stating missing macro layout, but in the Confluence page it isn’t a layout, it is several macros. Here is an example:
in XWiki:
Another annoyance (a small one), the Confluence sections info contents are imported, but in XWiki are coming with html tags, which is not very nice looking.
<p> documented in <ins>path/to/some/document</ins></p>
I can provide a pdf copy showing the Filter Streams Converter options used, eventually.
I will be grateful for advices and leads that can help me figure how to avoid missing contents, and the other stuff.
