Affects Version/s: 10.5
Fix Version/s: None
When a site is copied, any dependent resources should be copied too and links to the resource should be updated to point to the new site.
Currently we will need to process all html files in resources and URL resources. And then any HTML content that is in a tool.
Have a kernel util or service which given a string will ensure all resources referenced in that string are copied into the new site, returning the updated string.
As most HTML files should be reasonably small reading the whole file into memory isn't too serious. I think we should warn about processing files large than 1MB and refuse to process 10MB files.
Tools can then use this service to update their HTML and ensure that the resources exist in the new site.
We will only copy content from the source site, any references to content in other sites will remain untouched, otherwise content that is designed to be shared may well end up duplicated across sites.