Re: Row pattern recognition

> On Tue, Apr 23, 2024 at 8:13 PM Tatsuo Ishii <ishii(at)sraoss(dot)co(dot)jp> wrote:
>> SELECT v.a, count(*) OVER w
>> FROM (VALUES ('A'),('B'),('B'),('C')) AS v (a)
>> WINDOW w AS (
>> ORDER BY v.a
>> ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING
>> PATTERN (B+)
>> DEFINE B AS a = 'B'
>> )
>> a | count
>> ---+-------
>> A | 0
>> B | 2
>> B |
>> C | 0
>> (4 rows)
>>
>> Here row 3 is skipped because the pattern B matches row 2 and 3. In
>> this case I think cont(*) should return 0 rathern than NULL for row 3.
>
> I think returning zero would match Vik's explanation upthread [1],
> yes. Unfortunately I don't have a spec handy to double-check for
> myself right now.

Ok. I believe you and Vik are correct.
I am modifying the patch in this direction.
Attached is the regression diff after modifying the patch.

Best reagards,
--
Tatsuo Ishii
SRA OSS LLC
English: http://www.sraoss.co.jp/index_en/
Japanese:http://www.sraoss.co.jp