Chris
Chris
See https://github.com/recap-build/recap/issues/264 https://github.com/recap-build/recap/pull/411 For discussion about PG's ARRAY implementation.
Aw, dang. Yea, I can work around using the ORM for now. Thanks anyway. =) Do you happen to know if there's an issue opened on duckdb?
Interestingly, it appears SQLite behaves like duckdb-engine. I wonder if Postgres is behaving differently because it's using JSONB instead of JSON as the column type...
In case anyone else comes across this, you can use a query like this with the ORM: ``` rows = session.execute(select(CatalogEntry.metadata_).where( text(query) )).fetchall() or [] ``` To get around the...
Got it. Thanks for looking into this!
> Debezium (considering that it knows the precise column types) will have to zero-pad the values on its own (whenever the value is shorter then expected). Will this work, though?...
> If 1 then here couldn't we detect the difference in actual vs column length and right-pad the byte[]? But if it's 1, wouldn't you not know whether `a` was...
@shyiko if that's the case, I think it's actually a pretty fair argument to make DBZ assume that a fixed-length BINARY column should include the right padded 0's. Basically, my...
Might just be a wrapper on Paramiko. Maybe we can learn something from it, though.