Message-ID: <4E9A1738.6090108@wulczer.org>
Date: Sun, 16 Oct 2011 01:28:56 +0200
From: =?UTF-8?B?SmFuIFVyYmHFhHNraQ==?= <wulczer@wulczer.org>
User-Agent: Mozilla/5.0 (X11; Linux i686;
	rv:5.0) Gecko/20110624 Thunderbird/5.0
MIME-Version: 1.0
To: Postgres - Hackers <pgsql-hackers@postgresql.org>
Subject: plpython SPI cursors
Content-Type: multipart/mixed; boundary="------------060501080101020408060607"

This is a multi-part message in MIME format.
--------------060501080101020408060607
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit

Hi,

attached is a patch implementing the usage of SPI cursors in PL/Python.
Currently when trying to process a large table in PL/Python you have
slurp it all into memory (that's what plpy.execute does).

This patch allows reading the result set in smaller chunks, using a SPI
cursor behind the scenes.

Example usage:

cursor = plpy.cursor("select a, b from hugetable")
for row in cursor:
    plpy.info("a is %s and b is %s" % (row['a'], row['b']))

The patch itself is simple, but there's a lot of boilerplate dedicated
to opening a subtransaction and handling prepared plans. I'd like to do
some refactoring of they way PL/Python uses SPI to reduce the amount of
boilerplate needed, but that'll come as a separate patch (just before
the patch to split plpython.c in smaller chunks).

This feature has been sponsored by Nomao.

Cheers,
Jan

PS: I already added it to the November CF.

J

--------------060501080101020408060607
Content-Type: text/x-diff;
 name="0001-Add-cursor-support-to-plpythonu.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
	filename="0001-Add-cursor-support-to-plpythonu.patch"


--------------060501080101020408060607--