Migration tool from Atlassian Confluence to Nextcloud Collectives

Hi there !

hoping to get some ideally constructive :slight_smile: feedback on this thing I’ve built (well Claude did, I orchestrated and reviewed and in the process of bug fixing now).

similar to Migration from Confluence to Collectives topic some time ago, we had a need to migrate a whole bunch of stuff from Confluence which is to be shut down, over to Collectives.

after checking things out - it seemed to me that this should be somewhat script-able so after several iterations and re-runs from scratch I think we have something that mostly works - GitHub - realaaa/confluence-to-collectives

Sill a lot of niche issues etc to polish out and stomp those bugs, but it does the job !

I will be extending it over this week, and adding some more caveats etc - of course use at your own discretion, and if you see something wrong with it - please let me know !

Regards
Aleksey

1 Like

Dear @realaaa, thanks for sharing and great to see your tool! One of the Collectives maintainers here. I started working on an importer for Markdown files into Collectives myself, see Add support to import a markdown directory by mejo- · Pull Request #2196 · nextcloud/collectives · GitHub. It’s currently only tested with importing Markdown extracted from a Dokuwiki, but I’d love to see this extended.

From my experience so far, the biggest challenge is to adjust internal links and handle attachments. I see that your tool tries to handle the second, but from what I get, it doesn’t follow the convention to upload attachments to the .attachments.<fileId>/ directory next to the markdown file, right?

Besides, it was a bit of a challenge to ensure that the Markdown files have the file structure that Collectives expects. Pages with subpages need to be put into <pageTitle>/Readme.md and at the same time we have to ensure that there’s not both <pageTitle>.md and <pageTitle>/Readme.md to avoid conflicts.

My PR attempts to add the import functionality to Collectives itself, as we already have markdown parsing code here and Collectives knows about dealing with page name conflicts and alike.

Would you be interested in testing the import part with the code in my PR and adjust it gradually so it works with the Markdown generated by your exports from Confluence?

If you prefer to chat, feel invited to join our community chat here.

1 Like

thank you very much for that @mejo ! this looks awesome yes, and solves the second part of the problem perfectly yes

I will test it out during the next several days, and yes of course more than happy to switch that part to your Markdown Import tool, why re-invent that bicycle

then I can only focus on extracting from Confluence and converting to Markdown part :+1:

1 Like

Great, looking forward to your feedback @realaaa :folded_hands:

1 Like

hey hey @mejo Jonas thanks again for that suggestion !

got a bit side tracked, but took a closer look at this - and indeed for now I’ll just leave it as possible Issue for the future - Track: Nextcloud Collectives PR #2196 — server-side OCC markdown import · Issue #4 · realaaa/confluence-to-collectives · GitHub

as of course it would change the logic of the migration tool quite dramatically

I’ll keep watching your PR though - great work!!