Re: Binary search in ScalarArrayOpExpr for OR'd constant arrays - Mailing list pgsql-hackers

From James Coleman
Subject Re: Binary search in ScalarArrayOpExpr for OR'd constant arrays
Date
Msg-id CAAaqYe9ZPeFxw126uao4KgAXPO_Exr6t8zeNLSDcXdEf93m53A@mail.gmail.com
Whole thread Raw
In response to Re: Binary search in ScalarArrayOpExpr for OR'd constant arrays  (David Rowley <dgrowleyml@gmail.com>)
List pgsql-hackers
On Sun, Apr 26, 2020 at 11:44 PM David Rowley <dgrowleyml@gmail.com> wrote:
>
> On Mon, 27 Apr 2020 at 15:12, James Coleman <jtc331@gmail.com> wrote:
> > While working on this I noticed that dynahash.c line 499 has this assertion:
> >
> > Assert(info->entrysize >= info->keysize);
> >
> > Do you by any chance know why the entry would need to be larger than the key?
>
> Larger or equal. They'd be equal if you the key was the data, since
> you do need to store at least the key.  Looking at the code for
> examples where dynahash is used in that situation, I see
> _hash_finish_split().

Ah, I was thinking of it as key and value being separate sizes added
together rather than one including the other.

Thanks,
James



pgsql-hackers by date:

Previous
From: "Jonathan S. Katz"
Date:
Subject: Re: Poll: are people okay with function/operator table redesign?
Next
From: Tom Lane
Date:
Subject: Re: Poll: are people okay with function/operator table redesign?