On Sat, 21 Jan 2023 at 02:04, Ted Yu <yuzhih...@gmail.com> wrote: > > > > On Fri, Jan 20, 2023 at 4:51 PM Matthias van de Meent > <boekewurm+postg...@gmail.com> wrote: >> >> Attached a patch that solves this specific issue in a >> binary-compatible way. I'm not super happy about relying on behavior >> of callers of BeginCopyFrom (assuming that users that run copy >> concurrently will not provide a ParseState* to BeginCopyFrom), but it >> is what it is. > > Is it possible to check range_table / rteperminfos so that we don't introduce > the bool field ?
I think yes, but I'm not sure we can depend on rteperminfos to be set, and the same for p_rtable. I also don't think it's a good idea for code clarity: there is no good reason why the (un)availability of either range_table or rteperminfos would allow progress reporting - it would require additional extensive documentation around both the usage sites and the field itself. Adding a well-named field provides a much better experience in my opinion. If someone were opposed to adding that field in backbranches I'm fine with using one of these instead, assuming additional clear documentation is added as well. - Matthias
