From: Tom Lane <tgl@sss.pgh.pa.us>
To: Ivan Panchenko <i.panchenko@postgrespro.ru>
cc: pgsql-www@lists.postgresql.org
Subject: Re: Mailing list search engine: surprising missing results?
In-reply-to: <79b3eb6e-152e-3c56-7b71-51d091c0f6d9@postgrespro.ru>
References: 
 <CALDQ5NxzgeXHRCD4dS_6qz+nn01ivi3i1ZEtD2DmC779i0=iSQ@mail.gmail.com>
 <ab4184b7ab84623be10c4676e090cc27ae78b355.camel@cybertec.at>
 <Ye79wNIXsyhwwwce@momjian.us> <2150096.1643057249@sss.pgh.pa.us>
 <CAF4Au4yttKJ1KAP-cO+HMLQ2_66vmx0dLTBUbE4W8Aa64foafg@mail.gmail.com>
 <22d5245c9c5a9aa05a0510bdd52458812140a870.camel@cybertec.at>
 <2257661.1643127753@sss.pgh.pa.us>
 <79b3eb6e-152e-3c56-7b71-51d091c0f6d9@postgrespro.ru>
Comments: In-reply-to Ivan Panchenko <i.panchenko@postgrespro.ru>
	message dated "Tue, 25 Jan 2022 20:02:36 +0300"
MIME-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-ID: <2274254.1643133268.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Tue, 25 Jan 2022 12:54:28 -0500
Message-ID: <2274255.1643133268@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/2274255.1643133268%40sss.pgh.pa.us>
Precedence: bulk

Ivan Panchenko <i.panchenko@postgrespro.ru> writes:
> The actual explanation can be seen from comparing a tsvector with a tsqu=
ery.
> To avoid stemming effects, we use the simple configuration below.

> # select plainto_tsquery('simple','boyers-moore');

>  =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 plainto_ts=
query
> -------------------------------------
>  =C2=A0'boyers-moore' & 'boyers' & 'moore'

> # select to_tsvector('simple','boyers-moore-horspool');

>  =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 t=
o_tsvector
> -------------------------------------------------------------
>  =C2=A0'boyers':2 'boyers-moore-horspool':1 'horspool':4 'moore':3

> Obviously, such tsvector does not match the above tsquery. I think,a bet=
ter tsquery for this query would be

>  =C2=A0'boyers-moore' | ('boyers' & 'moore')

> May be, it is worth changing to_tsquery() behavior for such cases.

Changing the behavior of to_tsquery is certainly a lot less scary
than changing to_tsvector --- it wouldn't call the validity of
existing tsvector indexes into question.

I see that to_tsquery is even sillier than plainto_tsquery:

regression=3D# select to_tsquery('simple','boyers-moore');
               to_tsquery                =

-----------------------------------------
 'boyers-moore' <-> 'boyers' <-> 'moore'
(1 row)

which is absolutely not a sane translation.

It seems to me that in both cases we'd be better off generating
"'boyers' <-> 'moore'", without the compound token at all.
Maybe there's a case for the weaker 'boyers' & 'moore' translation,
but I think if people wanted that they'd just enter separate words.

			regards, tom lane