Since it incorporates fairly significant chunks of code from a number of places and is really a proof of concept rather than demo/example I don't think it can really be included as a TAP test module, not without more rewriting than I currently have time for anyway.
But it shows concept of DDL replication.
Yeah, a part of it anyway. As I outlined in another thread some time ago, getting DDL replication right is fairly tricky and requires more changes than just capturing the raw DDL (or even deparsed DDL) and sending it over the wire.
It's a useful proof of concept, but way too big to use as an in-tree regression for logical WAL messages as I was hoping.
