Re: plperl functions -- can they call each other?

Поиск
Список
Период
Сортировка
От ayhan
Тема Re: plperl functions -- can they call each other?
Дата
Msg-id 000701c2c84d$34be02e0$a000a8c0@burgaz
обсуждение исходный текст
Ответ на plperl functions -- can they call each other?  (will trillich <will@serensoft.com>)
Список pgsql-general

AFAIK, ==> the short answer is : No, not directly. As, apparently, the language handler just makes these functions anonymous subroutines.

 

There are 2 ways that you can deal with this though (not counting the possible SQL call via PgSPI).

 

A) If you are willing to switch to "plperlU" (the untrusted vesion) --- with all the SECURITY and STABILITY implications (see the doc for that), there is a way :

 

      1) Define the subroutine in a perl module

      2) Put the module in a directory that would be in the @INC.

      3) "Use" that module in your plperlu "procedure".     (which will load it just once per backend).

      4) And then just call the desired subroutine.

 

This has the added advantage that such subroutines are reusable not just within pgSQL but anything that you do with Perl. Besides, this way you wouldn't have to woory about the -very anoying- escaping backslashes and doubling single quotes problem.

 

This is the normal Perlish way.

 

B) The other way is a very hairy hack with the symbol table. I don't think you would even want to know.... In summary, this would mean actually creating an anonymous subroutine within your proc, assigning it a name in the symbol table if it hasn't been done already (by an earlier call of your proc). And then using that name in your other procs. Besides you would have to make sure that the code that magically adds the subroutine to the symbol table gets executed BEFORE any calls to that subroutine within any one session. UGLY. UGLY. UGLY.   -- only if you can't live with plperlu.

 

Hope this helps,

Ayhan

 

-----Original Message-----

From: pgsql-general-owner@postgresql.org [mailto:pgsql-general-owner@postgresql.org] On Behalf Of will trillich

Sent: mercredi 29 janvier 2003 20:52

To: pgsql-general@postgresql.org

Subject: [GENERAL] plperl functions -- can they call each other?

 

create or replace function to_partial_ymd(int2,int2,int2)returns int4 as ' my ($y,$m,$d) = @_; $m = 0 unless $y; $m = 0 unless 1 <= $m and $m <= 12; $d = 0 unless $m; $d = 0 unless (1 <= $d and

      (

            # 28 days valid for ANY month:

            ($d <= 28)

            or

            # 29 days valid if not february -- or in a leap year

            ($d <= 29 and ( $m != 2 or $y % 4 == 0 ))

            or

            # 30 days valid if not february

            ($d <= 30 and $m != 2)

            or

            # 31 days valid jan/mar/may/jul/aug/oct/dec

            ($d <= 31 and $m =~ /^(1|3|5|7|8|10|12)$/)

      )

);

return ($y << 16) + ($m << 8) + ($d << 0);

' language 'plperl'; -- '

 

sure would be nice to have such hefty day-of-month logic just ONCE, and be able to call it from other functions...

 

is there still no way for one plperl function to call another plperl function? (i'm hoping that section 25.3.4 is out-of-date...)

 

--

There are 10 kinds of people:

ones that get binary, and ones that don't.

 

will@serensoft.com

http://sourceforge.net/projects/newbiedoc -- we need your brain! http://www.dontUthink.com/ -- your brain needs us!

 

Looking for a firewall? Do you think smoothwall sucks? You're probably right... Try the folks at http://clarkconnect.org/ !

 

---------------------------(end of broadcast)---------------------------

TIP 4: Don't 'kill -9' the postmaster

 

 

В списке pgsql-general по дате отправления:

Предыдущее
От: Robert Treat
Дата:
Сообщение: Re: how do I get a log of SQL executed by server?
Следующее
От: charette@writeme.com (Stephane Charette)
Дата:
Сообщение: Cannot insert a duplicate key into unique index pg_calss_oid_index