Re: Incorrect information about GIN-index in RUM's docs

On 15.11.2016 11:12, Andreas Joseph Krogh wrote:
> På tirsdag 15. november 2016 kl. 08:48:37, skrev Oleg Bartunov
> <obartunov@gmail.com <mailto:obartunov@gmail.com>>:
>
>
>
>     On Tue, Nov 15, 2016 at 10:37 AM, Andreas Joseph Krogh
>     <andreas@visena.com <mailto:andreas@visena.com>> wrote:
>
>         This item isn't valid, is it?
>
>
>           * There isn't phrase search with |GIN| index. This problem
>             relates with previous problem. It is need position
>             information to perform phrase search.
>
>         RUM being an extention, and having index-access-methods as
>         extentions is new in 9.6, where phrase-search was introduced,
>         this isn't really true, or am I missing something?
>
>
>
>     RUM is very good for phrase search, since it has access to
>     coordinates in index, check slides from PGconf.eu
>     (http://www.sai.msu.su/~megera/postgres/talks/pgconfeu-fts-2016.pdf).
>     Where did you find the citation ?
>
>
> https://github.com/postgrespro/rum#user-content-introduction
>
> I assume I read the item wrong tho. As it's phrased, it (for me at least
> but I'm Norwegian) seems like phrase-search is not possible with GIN,
> which it definitely is. So I assume the real issue, that this item is
> trying to point out, is that phrase-search isn't really /optimized/ with
> GIN, just syntactic sugar. And with RUM it is optimized, not just
> syntactic sugar. Am I correct assuming this?

Yes. I suppose the original explanation about phrase search in GIN is
confusing and ambiguous. GIN supports it, but it requires additional
bitmap heap scan and so it slower.

I fixed it. Thank you.

>
> --
> *Andreas Joseph Krogh*
> CTO / Partner - Visena AS
> Mobile: +47 909 56 963
> andreas@visena.com <mailto:andreas@visena.com>
> www.visena.com <https://www.visena.com>
> <https://www.visena.com>
>


--
Artur Zakirov
Postgres Professional: http://www.postgrespro.com
Russian Postgres Company