On Tue, Mar 2, 2021, at 06:31, Tom Lane wrote: > "Joel Jacobson" <[email protected]> writes: > > Unless fixed, then the way I see it, I don't think we can use int4range[] > > for regexp_positions(), > > Yeah. It's a cute idea, but the semantics aren't quite right.
Having abandoned the cute idea that didn't work,
here comes a new patch with a regexp_positions() instead returning
setof record (start_pos integer[], end_pos integer[]).
Example:
SELECT * FROM regexp_positions('foobarbequebazilbarfbonk',
$re$(b[^b]+)(b[^b]+)$re$, 'g');
start_pos | end_pos
-----------+---------
{3,6} | {6,11}
{11,16} | {16,20}
(2 rows)
Based on HEAD (040af779382e8e4797242c49b93a5a8f9b79c370).
I've updated docs and tests.
/Joel
0002-regexp-positions.patch
Description: Binary data
