Same logic is applicable, in the event of failure false or invalid
resource maybe returned.
Ilia
Markus Bertheau wrote:
Ð ÐÐÐ, 08.11.2004, Ð 16:33, Ilia Alshanetsky ÐÐÑÐÑ:
The error functions do so because they are likely to be called if
creation of a connection fails so there may not always be
Ð ÐÐÐ, 08.11.2004, Ð 16:33, Ilia Alshanetsky ÐÐÑÐÑ:
> The error functions do so because they are likely to be called if
> creation of a connection fails so there may not always be a valid
> resource. The state checking function would presumably be only run on a
> valid connection and consequentl
The error functions do so because they are likely to be called if
creation of a connection fails so there may not always be a valid
resource. The state checking function would presumably be only run on a
valid connection and consequently should be strict about the resource
parameter (IMO).
Ili
Ð ÐÐÐ, 08.11.2004, Ð 16:04, John Coggeshall ÐÐÑÐÑ:
> if (zend_parse_parameters_ex(ZEND_PARSE_PARAMS_QUIET, ZEND_NUM_ARGS()
> TSRMLS_CC, "r" ..
>
> why would you silence the error for the parameter when this function
> requires a valid Postgres resource to do anything?
I copied the behaviour
if (zend_parse_parameters_ex(ZEND_PARSE_PARAMS_QUIET, ZEND_NUM_ARGS()
TSRMLS_CC, "r" ..
why would you silence the error for the parameter when this function
requires a valid Postgres resource to do anything?
John
On Mon, 2004-11-08 at 07:44, Markus Bertheau wrote:
> Hi,
>
> attached is a p
Hi,
attached is a patch which implements a function pg_result_sqlstate. This
function is needed because the error messages returned by
pg_result_error is localized and can therefor not be parsed reliably.
The patch is against PHP5. Barring any objections please apply the patch
for PHP 5.0.3.
Tha
