Parameters don't line up with the names properly about plrust

[eeeebbbbrrrr]

Woah.

Thanks for the report.

[workingjubilee]

fwiw this test would be written the same as

CREATE OR REPLACE FUNCTION test(a int, b int) 
  RETURNS int AS 
$$
  a
$$ LANGUAGE plrust;

I have reproduced the error as a test and while my initial hypothesis was a problem in pgx, it seems to be unique to PL/Rust and the equivalent test in pgx passes fine.

[workingjubilee]

...This appears to be a "Heisenbug", which may explain why we have not seen it before. That is, it does not seem to happen consistently, and the ordering that happens somehow can change when the functions are rebuilt and retested. It also seems to be exacerbated by simultaneously submitting multiple PL/Rust functions at once, but is not exclusive to that.

[workingjubilee]

Ah. The lesson here is to never use HashMap (which is, by default, HashMap<K, V, RandomState>... i.e. the data is hashed with a random state seed selected upon creation of the HashMap) and expect a deterministic ordering.