[GENERAL] textsubstr() ...? for postgres 7 beta5

[nathan]

I have installed the beta 7 for postgres but... it seems to be missing
some text functions ? l

like the textsubstr ?

Is this beacuse it is a "beta" and the final release will include these
or are they forever droped ?

Please let me know

thank you

nathan

[GENERAL] New eMail address

[Jan Wieck]

Hi folks,

would anyone please update his aliases?  My new email address
is

[EMAIL PROTECTED]

I'll be able to read mail addressed to [EMAIL PROTECTED]  until
end  of  May.  Mail  reaching  that  address  after  May will
probably get lost.


Jan

-- 

#==#
# It's easier to get forgiveness for being wrong than for being right. #
# Let's break this rule - forgive me.  #
#== [EMAIL PROTECTED] #

[GENERAL] Re: [SQL] textsubstr() ...? for postgres 7 beta5

[Tom Lane]

[EMAIL PROTECTED] writes:
> I have installed the beta 7 for postgres but... it seems to be missing
> some text functions ? l

> like the textsubstr ?

Huh?  There is no function named textsubstr() in 6.5.*, either.  Perhaps
you had some locally user-defined functions you forgot to reinstall?

regards, tom lane

[GENERAL] implementing refcounts across tables.

[Alfred Perlstein]

Think of this problem as handling unix style filesystem hardlinks.

Consider two tables:

create table link (
  data_id int4,
  link_date timestamp
);

create table data (
  date_id int4 PRIMARY KEY,
  data_txt text
);

We may have multiple "link" rows pointing at the same data.

What I would like to implement is a rule that when a row is deleted
from "link" then if no other rows in "link" reference a row in "data"
then the row in data is deleted as well.

Here's what I've tried:

create rule
  cascade_clean_data
as on delete to link
do delete from data
where
  data_id = OLD.data_id 
  AND '1' = (select count(*) from link where data_id = OLD.data_id)
;

I've also tried '0' for the second conditional, but niether seem to
work.

So, is there a way to accomplish this automatic refcounting either
with rules or some other trick?

Also, I think the count(*) is a bad idea because we only need to know
if a single entry besideds the one we are deleteing exists, not the
actual count.

Any suggestions?

thanks,
-- 
-Alfred Perlstein - [[EMAIL PROTECTED]|[EMAIL PROTECTED]]
"I have the heart of a child; I keep it in a jar on my desk."

Re: [GENERAL] Re: [SQL] textsubstr() ...? for postgres 7 beta5

[nathan]

Tom Lane wrote:

> [EMAIL PROTECTED] writes:
> > I have installed the beta 7 for postgres but... it seems to be missing
> > some text functions ? l
>
> > like the textsubstr ?
>
> Huh?  There is no function named textsubstr() in 6.5.*, either.  Perhaps
> you had some locally user-defined functions you forgot to reinstall?
>
> regards, tom lane

Opps...

Ok, well I made a mistake

it is not textsubstr() but this : (found in  6.5.2 redhat )

result  |function |arguments  |description
+-+---+---

text|text_substr  |text int4 int4 |return portion of string

Is this in 7.0 ?

(PS this is not user defined, well at least I did not make  it ...? )

thank you

nathan

Re: [HACKERS] Re: [GENERAL] Re: [SQL] textsubstr() ...? for postgres 7 beta5

[Tom Lane]

[EMAIL PROTECTED] writes:
> it is not textsubstr() but this : (found in  6.5.2 redhat )

> result  |function |arguments  |description
> +-+---+--
> text|text_substr  |text int4 int4 |return portion of string

> Is this in 7.0 ?

Looks like it's called just plain "substr" now.  See
http://www.postgresql.org/docs/postgres/functions.htm

regards, tom lane