Message-ID: <9138745.post@talk.nabble.com>
Date: Sat, 24 Feb 2007 15:04:51 -0800 (PST)
From: chrisj <chrisj.wood@sympatico.ca>
To: pgsql-sql@postgresql.org
Subject: Re: can someone explain confusing array indexing nomenclature
In-Reply-To: <200702200942.18500.achill@matrix.gatewaynet.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
References: <8971770.post@talk.nabble.com>
	<200702151224.12455.achill@matrix.gatewaynet.com>
	<8989242.post@talk.nabble.com>
	<200702161034.43399.achill@matrix.gatewaynet.com>
	<9009934.post@talk.nabble.com>
	<200702200942.18500.achill@matrix.gatewaynet.com>


I guess you could say the [2] is discarded since the value "2" is at the to=
p
or beyond the top of the range.

Achilleas Mantzios wrote:
>=20
> =CE=A3=CF=84=CE=B9=CF=82 =CE=A0=CE=B1=CF=81=CE=B1=CF=83=CE=BA=CE=B5=CF=85=
=CE=AE 16 =CE=A6=CE=B5=CE=B2=CF=81=CE=BF=CF=85=CE=AC=CF=81=CE=B9=CE=BF=CF=
=82 2007 20:35, =CE=BF/=CE=B7 chrisj =CE=AD=CE=B3=CF=81=CE=B1=CF=88=CE=B5:
>> I am quite sure the [2] is not discarded, easy enough to test but I don'=
t
>> have access to PG at the moment.
>=20
> Well it should, since
>=20
> dynacom=3D# SELECT=20
> (CAST( '{{meeting,lunch},{training,presentation}}' as text[][]))[1:1];
>        text
> -------------------
>  {{meeting,lunch}}
> (1 row)
>=20
> dynacom=3D# SELECT=20
> (CAST( '{{meeting,lunch},{training,presentation}}' as text[][]))[1:1][1];
>     text
> -------------
>  {{meeting}}
> (1 row)
>=20
> dynacom=3D# SELECT=20
> (CAST( '{{meeting,lunch},{training,presentation}}' as text[][]))[1:1][2];
>        text
> -------------------
>  {{meeting,lunch}}
> (1 row)
>=20
> dynacom=3D# SELECT=20
> (CAST( '{{meeting,lunch},{training,presentation}}' as text[][]))[1:1][3];
>        text
> -------------------
>  {{meeting,lunch}}
> (1 row)
>=20
> dynacom=3D# SELECT=20
> (CAST( '{{meeting,lunch},{training,presentation}}' as
> text[][]))[1:1][1000];
>        text
> -------------------
>  {{meeting,lunch}}
> (1 row)
>=20
> dynacom=3D#                           =20
>=20
>>
>> Achilleas Mantzios wrote:
>> > =CE=A3=CF=84=CE=B9=CF=82 =CE=A0=CE=AD=CE=BC=CF=80=CF=84=CE=B7 15 =CE=
=A6=CE=B5=CE=B2=CF=81=CE=BF=CF=85=CE=AC=CF=81=CE=B9=CE=BF=CF=82 2007 18:55,=
 =CE=BF/=CE=B7 chrisj =CE=AD=CE=B3=CF=81=CE=B1=CF=88=CE=B5:
>> >> Thanks Achilleas,
>> >>
>> >> I see what you are saying, but if we consider just the index "[2]" fo=
r
>> a
>> >> moment,
>> >> it means something different depending upon the context  (in one case
>> it
>> >> means "2" and in the other case it means "1:2") and the context is
>> >> determined by the format of indexes on other dimensions.
>> >>
>> >> I believe I understand....but incredibly confusing.
>> >
>> > Now that i think about it again, i speculate that the [2] is discarded=
.
>> >
>> >> - chris
>> >>
>> >> Achilleas Mantzios wrote:
>> >> > =CE=9E=C2=A3=CE=9F=C2=84=CE=9E=CE=89=CE=9F=C2=82 =CE=9E=EF=BF=BD=CE=
=9E=CE=85=CE=9F=C2=84=CE=9E=C2=AC=CE=9F=C2=81=CE=9F=C2=84=CE=9E=C2=B7 14 =
=CE=9E=C2=A6=CE=9E=CE=85=CE=9E=C2=B2=CE=9F=C2=81=CE=9E=CE=8F=CE=9F=C2=85=CE=
=9E=C2=AC=CE=9F=C2=81=CE=9E=CE=89=CE=9E=CE=8F=CE=9F=C2=82 2007 21:31, =CE=
=9E=CE=8F/=CE=9E=C2=B7
>> >>
>> >> chrisj
>> >
>> > =CE=9E=C2=AD=CE=9E=C2=B3=CE=9F=C2=81=CE=9E=C2=B1=CE=9F=C2=88=CE=9E=CE=
=85:
>> >> >> given the following table:
>> >> >>
>> >> >> protocal2=3D> select * from sal_emp ;
>> >> >>  name  |      pay_by_quarter       |                 schedule
>> >>
>> >>
>> -------+---------------------------+------------------------------------
>> >>
>> >> >>--- ---- Bill  | {10000,10000,10000,10000} |
>> >> >> {{meeting,lunch},{training,presentation}}
>> >> >>  Carol | {20000,25000,25000,25000} |
>> >> >> {{breakfast,consulting},{meeting,lunch}}
>> >> >> (2 rows)
>> >> >>
>> >> >> why do the following two queries yield different results??
>> >> >>
>> >> >> protocal2=3D> SELECT schedule[1][2] FROM sal_emp WHERE name =3D 'B=
ill';
>> >> >>  schedule
>> >> >> ----------
>> >> >>  lunch
>> >> >> (1 row)
>> >> >>
>> >> >> protocal2=3D> SELECT schedule[1:1][2] FROM sal_emp WHERE name =3D
>> 'Bill';
>> >> >>      schedule
>> >> >> -------------------
>> >> >>  {{meeting,lunch}}
>> >> >> (1 row)
>> >> >
>> >> > The [n:m] notation denotes a slice of the array (not element).
>> >> > So schedule[1][2] is the Array element on 2nd col of 1st row,
>> >> > while schedule[1:1][2] could mean
>> >> > the second row of the subarray schedule[1:1][1:2].
>> >> > So these two are foundamentally different things.
>> >> > In my 7.4 even if you gave
>> >> > SELECT schedule[1:1][888] FROM sal_emp WHERE name =3D 'Bill';
>> >> > you would still get  {{meeting,lunch}} as a result.
>> >> > (Right or wrong is another story).
>> >> > Anyway the first time you query for a "text",
>> >> > the second time you query for a "text[]", so you should expect
>> >> > different results.
>> >> > --
>> >> > Achilleas Mantzios
>> >> >
>> >> > ---------------------------(end of
>> >>
>> >> broadcast)---------------------------
>> >>
>> >> > TIP 3: Have you checked our extensive FAQ?
>> >> >
>> >> >                http://www.postgresql.org/docs/faq
>> >
>> > --
>> > Achilleas Mantzios
>> >
>> > ---------------------------(end of
>> broadcast)---------------------------
>> > TIP 3: Have you checked our extensive FAQ?
>> >
>> >                http://www.postgresql.org/docs/faq
>=20
> --=20
> Achilleas Mantzios
>=20
> ---------------------------(end of broadcast)---------------------------
> TIP 7: You can help support the PostgreSQL project by donating at
>=20
>                 http://www.postgresql.org/about/donate
>=20
>=20

--=20
View this message in context: http://www.nabble.com/can-someone-explain-con=
fusing-array-indexing-nomenclature-tf3229165.html#a9138745
Sent from the PostgreSQL - sql mailing list archive at Nabble.com.