Michael Banck wrote: > Hi, > > Am Mittwoch, den 21.03.2018, 09:54 -0300 schrieb Alvaro Herrera: > > Michael Banck wrote: > > > I apparently managed to screw up so badly that no PQerrorMessage was > > > set, so saw the above (which indeed has no error message after the > > > colon). > > > > Well, maybe that's a different bug, then: maybe we should print > > something other than PQerrorMessage (or maybe PQerrorMessage should not > > return empty!). Can you reproduce the problem? > > It was while I was hacking on the code and mistyped something, so I > think this is entirely my fault and not a valid concern. From what I > could tell, I got PGRES_FATAL_ERROR back (I printed the int value, which > was 7), so that could explain why there was no PQerrorMessage?
If the PGresult passed to PQerrorMessage was NULL, then there's nowhere to stuff a fabricated error message either, so it all sounds good to me. -- Álvaro Herrera https://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
