Older versions of (Windows) Funnelback (v15 and earlier) were able to index content from content manager (and earlier RM/TRIM versions) but this had a lot of requirements/caveats and would not have worked in a SaaS environment.
This was highly complicated because the solution needed to support document level security (namely the search results are filtered so that the user running the search can only see items they have access to in the search results).
For this we needed:
* Funnelback running on Windows, installed on a server that was part of the Windows domain that hosted the CM server.
* Special configuration within the network to facilitate the authentication (including integration with AD, delegation rights for the server and configuration to allow Funnelback to do user impersonation).
* Special configuration and plugins inside Funnelback that enabled it to index the permission, detect the user's security group memberships and ultimately determine if the user was able to access the results.
This scenario is completely impossible to achieve in the Squiz DXP using the mechanisms that were supported on the Windows version of Funnelback.
There may be other ways of indexing the content from CM using the DXP integrations service (Squiz Connect), but for this to work you would need to be able to index the CM content, accessing it from the DXP. Furthermore, if you were wanting security across the results then there would be significant work that you would need to do to facilitate this (if it's even possible to get the security information you need via APIs etc.)
There are possibly also things you can do from the CM side (though I'm not at all familiar with what you can do). For example it might be possible to get CM to submit changes/content/etc. to a Funnelback push data source whenever a change is made in the system. In this sort of scenario FB becomes only responsible for indexing the content you give it, it's up to you to ensure the updates and all the relevant content is submitted to the index. And you still would have the problem of handling any security if document level security was a requirement.
I hope some of this is helpful in at least understanding the complexity of the problem!
