• Remove the deprecated and obsolete antora2plainasciidoc script and build step
• Add in upstream as much as possible from the downstream logic (Antora Assembler)

Use Antora Assembler to produce a monolithic AsciiDoc file for each module.

This technique is used to produce the downstream documentation.

Applying the same technique upstream will help prevent committing incompatible content.

See: gitlab.com/antora/antora-assembler

As a side effect, the Antora Assembler is producing PDF output. We have to be cautious with HTML specific content (use ifndef). I fixed legit vale warnings. I removed obsolete statements that were disabling vale, but are not required anymore. One is left, and I found a workaround to avoid an error during PDF generation. Discussion in Antora community about this workaround, see: antora.zulipchat.com/#narrow/stream/282400-users/topic/Assembler.20specific.20attributes/near/308605574

In downstream, we also have different steps that are not reproduced here:

Define different AsciiDoc attributes in the Antora playbook
Define different examples in an Antora Distributed Component Version
Remove some pages from the navigation by using an Antora extension that injects Antora Collector configuration.