Robert Haas writes:
> I feel like we ought to be doing this in a way where the output is
> properly escaped.
This is one of several reasons why code shouldn't be trying to scrape
the names out of the human-readable message.
regards, tom lane
--
Sent via pgsql-hackers ma
Robert Haas wrote:
> I feel like we ought to be doing this in a way where the output is
> properly escaped. Right now:
Yeah, that's been an occasional irritation for me, even though I
don't actually use quote characters in any of my quoted identifiers.
Also, tab-completion is irritating in no
On Tue, Jan 12, 2010 at 5:10 AM, Takahiro Itagaki
wrote:
> Andreas Joseph Krogh wrote:
>
>> ERROR: null value in column "created" violates not-null constraint
>
> It is easy to add the table name to the message, but ...
>
>> ERROR: null value in column "public"."mytable"."created" violates not-nu
Takahiro Itagaki writes:
> Do we have any guideline about the message for identifier names?
The issue has come up before. I think that what we really need here is
to implement additional fields in error message reports, so that the
name and schema name of the relevant object can be pulled out of
On Tuesday 12. January 2010 11.10.09 Takahiro Itagaki wrote:
>
> Andreas Joseph Krogh wrote:
>
> > ERROR: null value in column "created" violates not-null constraint
>
> It is easy to add the table name to the message, but ...
>
> > ERROR: null value in column "public"."mytable"."created" viol
Andreas Joseph Krogh wrote:
> ERROR: null value in column "created" violates not-null constraint
It is easy to add the table name to the message, but ...
> ERROR: null value in column "public"."mytable"."created" violates not-null
> constraint
> Oracle does this btw...
Do we have any guideli
