Community Announcements have moved! To stay up to date, please join the new Community Announcements group today. Learn more
×We use custom css classes on our pages for our PDF export. In August 2024 we moved from Confluence server to Confluence cloud and chose to use Storage Format Editor for Confluence FREE by K15t Labs. This solution works great.
However, with the forced updated of our legacy pages, the custom css is lost. We have no workaround.
It looks like the new editor not only strips existing css from legacy pages on conversion but also restrips any newly added css on every save.
This is a nightmare. Does anyone have any solutions? I'm also reaching out to K15t labs.
@Darryl Lee unfortunately, what @Thomas said is pretty close to what I received back from my K15t labs ticket, though Thomas did go into a bit more details.
Ugh, yeah, makes sense that the problem is the dreaded ADF.
A while back I ran into a "fun" bug where:
Turns out that ADF can "handle" merged cells just fine, and obviously it can display them in Jira just fine.
But Jira's own APIs to display content as rendered HTML will refuse to render the table.
Their "excuse" is that merged cells are not supported in Wiki Markup.
But we not trying to display Wiki Markup. Turns out Atlassian's own Word exporter, as well as all of the third-party tools, are asking for renderedBody, which is HTML, and HTML which 100% supports merged cells.
(And in fact, ADF format for merged cells uses the HTML attributes, colspan and rowspan!)
Anywho, yeah, Atlassian's own support for ADF kind of sucks, so go figure they make it hard for third-parties. I commented on this bug about it:
Hey everyone,
I'm one of the developers that work on the Storage Format Editor at K15t so I can give some details on the issue.
This issue is likely due to Atlassian shifting away from Storage Format and towards ADF for their editor and page rendering. In ADF custom CSS is not directly supported, only a few very limited styling options can be stored (pretty much the ones that the new Confluence editor already provides).
Confluence seems to still store a storage format representation of the page when using an storage format editor such as ours but since the new Confluence editor seems to exclusively work with ADF it will lose any information that can't be represented in ADF on the next page edit.
There's unfortunately nothing we can do about this on our end as the Confluence editor is the problem here. You could reach out to Atlassian support about this but I don't know if they will do anything about this since this implementation was likely an intentional choice on their side to reduce the amount of complexity in the Confluence page storage medium.
Cheers,
Thomas