Shouldn't psql -1 imply ON_ERROR_STOP?

On Saturday, July 25, 2009, Peter Eisentraut <peter_e@gmx.net> wrote:

When you run a file with psql -1/--single-transaction, and a command fails,
you get bombarded with

ERROR:  current transaction is aborted, commands ignored until end of
transaction block

for the rest of the file.

That would certainly be useful.

Personally I'd prefer it to default to that always, and not just in
-1, but that would break way too many old things I'm afraid...

/Magnus

Shouldn't -1 imply ON_ERROR_STOP or some variant by default?

--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

--
Magnus Hagander
Self: http://www.hagander.net/
Work: http://www.redpill-linpro.com/

On Sat, Jul 25, 2009 at 9:06 AM, Magnus Hagander<magnus@hagander.net> wrote:

On Saturday, July 25, 2009, Peter Eisentraut <peter_e@gmx.net> wrote:

When you run a file with psql -1/--single-transaction, and a command fails,
you get bombarded with

ERROR:  current transaction is aborted, commands ignored until end of
transaction block

for the rest of the file.

That would certainly be useful.

Personally I'd prefer it to default to that always, and not just in
-1, but that would break way too many old things I'm afraid...

Doing it always would be really annoying. I often reload dumps that
fail the grant statements but otherwise work. Admittedly, if I
planned ahead, I could avoid having the grants be present in the
dumps, but that would require planning ahead...

But +1 for doing it when -1 is used.

...Robert

Am 25.07.2009 um 15:00 schrieb Peter Eisentraut:

When you run a file with psql -1/--single-transaction, and a command
fails,
you get bombarded with

ERROR: current transaction is aborted, commands ignored until end of
transaction block

for the rest of the file.

Shouldn't -1 imply ON_ERROR_STOP or some variant by default?

Sounds reasonable, +1 from me.

Regards
Michael Paesold

--On Samstag, Juli 25, 2009 16:00:18 +0300 Peter Eisentraut
<peter_e@gmx.net> wrote:

When you run a file with psql -1/--single-transaction, and a command
fails, you get bombarded with

ERROR: current transaction is aborted, commands ignored until end of
transaction block

for the rest of the file.

Shouldn't -1 imply ON_ERROR_STOP or some variant by default?

Only if it could ensured that embedded SAVEPOINTS can be handled
properly...a quick check shows that ON_ERROR_STOP will stop any script even
when the errorneous command is probably rolled back by a subsequent
ROLLBACK TO:

SELECT 1;

SAVEPOINT A;

SELECT et; <-- ON_ERROR_STOP stops here

ROLLBACK TO A;

SELECT 2;

It seems -1 needs some smarter variant of ON_ERROR_STOP.

--
Thanks

Bernd