pl/python SPI in subtransactions - Mailing list pgsql-hackers

Here's a patch implementing a executing SPI in an subtransaction
mentioned in
http://archives.postgresql.org/pgsql-hackers/2010-12/msg01991.php. It's
an incremental patch on top of the plpython-refactor patch sent eariler.

Git branch for this patch:
https://github.com/wulczer/postgres/tree/spi-in-subxacts.

Without it the error handling in PL/Python is really broken, as we jump
between from a saught longjmp back into Python without any cleanup. As
an additional bonus you no longer get all the ugly "unrecognized error
in PLy_spi_execute_query" errors.

Cheers,
Jan