Absolute Cloud URLs are not usable in Server when migrating with simple domain name replacement

Summary

When migrating from Confluence Cloud to Confluence Server, if there are absolute URLs used in page links these are no usable in Confluence Server.

Usually we follow

• How Do I replace all 'hard' links to my old Base URL

to change the domain name, however the URLs are now completely different and there is no 1-to-1 mapping.

Steps to Reproduce

1. Set up a Confluence Cloud environment
2. Create 2 pages, PageA and PageB with some dummy content
3. Edit PageA, highlight some text and press Ctrl+K
4. Choose Web Link and for Address, use the full URL for PageB
5. Edit PageB, highlight some text and press Ctrl+K
6. Choose Web Link and for Address, use the full URL for PageA
7. Go to Backup Manager wiki/plugins/servlet/ondemandbackupmanager/admin and click Create Backup
8. Download Site_Backup.zip
9. Follow the steps in How Do I replace all 'hard' links to my old Base URL (note, this KB has been updated to reflect this bug) to change all the domain names from mycompany.atlassian.net to confluence.mycompany.com
10. Set up Confluence 6.6.1 and on the page Load Content select Restore From Backup
11. Choose the file Site_Backup.zip and proceed to set up Confluence as usual

Remember to start Confluence with

• Restore Passwords To Recover Admin User Rights

Expected Results

The URLs for the links between pages work as expected.

Actual Results

Confluence Cloud links are not 1-to-1 mappable to Confluence Server.

Notes

URLs in Confluence Cloud have the format of

• https://mycompany.atlassian.net/wiki/spaces/BUG/pages/229377/PageA

Whereas the similar link in Confluence Server would be either

• http://confluence.mycompany.com/display/BUG/PageA
• http://confluence.mycompany.com/pages/viewpage.action?pageId=229377

Workaround

The contentid is in the Cloud Confluence URL, so that can be used to map to a Confluence Server URL with /pages/viewpage.action?pageId=

For example, a regex to match all URLs would be

mycompany.atlassian.net/wiki/spaces/[A-Z0-9]+/pages/([0-9]+)/[A-Za-z0-9+.]+

and replace with (depending on regex engine)

confluence.mycompany.com/pages/viewpage.action?pageId=\1

For example if you have

You can use this sed script on entities.xml from the export

sed -r -e 's/mycompany\.atlassian\.net\/wiki\/spaces\/[A-Z0-9]+\/pages\/([0-9]+)\/[A-Za-z0-9+.]+/confluence.mycompany.com\/pages\/viewpage.action?pageId=\1/g' entities.xml