(v4) Redo CleanContent

[csmir]

As already addressed by several issues in the past, such as #2118 and #2317; CleanContent implementation has not been pulled off elegantly. The original implementor has not provided feedback on the matter, so it would be wise if this is picked up as a plan for v4.

Let us consider the following solutions:

• Move it to an extension package that implements a package to do this for us.
• Implement the package into the core library.
• Remove CleanContent entirely.
• Write a full implementation by ourselves.

The first idea would be preferred, considering the same would be done in #2589