On Thu, Sep 2, 2021, at 00:03, Daniel Gustafsson wrote:
> I can see value in a function like this one, and the API is AFAICT fairly
> aligned with what I as a user would expect it to be given what we already have.
Good to hear and thanks for looking at this patch.
I've fixed the problem due to the recent change in setup_regexp_matches(),
which added a new int parameter "start_search".
I pass 0 as start_search, which I think should give the same behaviour as before.
I also changed the assigned oid values in pg_proc.dat for the two new regexp_positions() catalog functions.
$ make check
=======================
All 209 tests passed.
=======================
/Joel