Re: Improving information_schema._pg_expandarray()

2024-05-13 Thread Dagfinn Ilmari Mannsåker
[ I got distracted while writing this follow-up and only just found it in my list of unsent Gnus buffers, and now it's probably too late to make it for 17, but here it is anyway while I remember. ] Tom Lane writes: > I happened to notice that information_schema._pg_expandarray(), > which has

Re: Improving information_schema._pg_expandarray()

2023-12-23 Thread Pavel Stehule
so 23. 12. 2023 v 19:18 odesílatel Tom Lane napsal: > I happened to notice that information_schema._pg_expandarray(), > which has the nigh-unreadable definition > > AS 'select $1[s], > s operator(pg_catalog.-) pg_catalog.array_lower($1,1) > operator(pg_catalog.+) 1 > from pg_c

Improving information_schema._pg_expandarray()

2023-12-23 Thread Tom Lane
I happened to notice that information_schema._pg_expandarray(), which has the nigh-unreadable definition AS 'select $1[s], s operator(pg_catalog.-) pg_catalog.array_lower($1,1) operator(pg_catalog.+) 1 from pg_catalog.generate_series(pg_catalog.array_lower($1,1),