After looking into error messages further, I see that the constraint name is
not included in NOT NULL or UNIQUE constraint errors although it is there in
CHECK constraint errors. The table.column that caused the error is in the
messages so it is possible to identify which column caused the error.
On Thu, May 8, 2014 at 10:52 PM, Walter Hurry wrote:
> On Thu, 08 May 2014 13:15:54 +0300, Constantine Yannakopoulos wrote:
> most if not all other databases do not implement this
>
> $ psql
> psql (9.3.4)
> Type "help" for help.
>
>
>
I meant that they do not implement a way to *directly* g
On Thu, 08 May 2014 13:15:54 +0300, Constantine Yannakopoulos wrote:
most if not all other databases do not implement this
$ psql
psql (9.3.4)
Type "help" for help.
misc=# create table parent
misc-# (a_id varchar(1),
misc(# constraint pk_parent primary key(a_id));
CREATE TABLE
misc=# create tab
On Thu, May 8, 2014 at 1:08 PM, Simon Slavin wrote:
> Perhaps a future version of sqlite3_extended_result_codes() can return a
> string as well as the code, the string being the name of the constraint
> that was violated.
>
Yes, it would be great to be able to get the necessary info in a structu
On 8 May 2014, at 10:40am, Constantine Yannakopoulos
wrote:
> On Thu, May 8, 2014 at 11:50 AM, Simon Slavin wrote:
>
>> That's your job. You're the programmer. SQLite does not talk to users
>> and its result values should never be shown to a user.
>
> Fair enough, but SQLite needs to give
On Thu, May 8, 2014 at 11:50 AM, Simon Slavin wrote:
> That's your job. You're the programmer. SQLite does not talk to users
> and its result values should never be shown to a user.
>
Fair enough, but SQLite needs to give me what I need to be able to
communicate something helpful to the user.
On 8 May 2014, at 9:00am, Constantine Yannakopoulos
wrote:
> When the fk violation occurs inside a large transaction,
> e.g. during a bulk data import it is essential that the user is given any
> possible help to be able to locate the data that violates ref integrity and
> fix it.
That's your
On May 8, 2014 4:39 AM, "phaworth" wrote:
>
> I'd vote for having this as a future enhancement under the control of a
> pragma or some other way of making it optional. Some of my tables have
more
> than 1 foreign key and without the constraint name I have to write
> application code to pre-check
I'd vote for having this as a future enhancement under the control of a
pragma or some other way of making it optional. Some of my tables have more
than 1 foreign key and without the constraint name I have to write
application code to pre-check for foreign key errors since I can't translate
the er
On Tue, May 6, 2014 at 11:17 PM, Richard Hipp wrote:
> On Tue, May 6, 2014 at 5:12 PM, Peter Haworth wrote:
>> It seems that foreign key errors on columns where the foreign key
>> definition has a constraint name don't include the constraint name in the
>> error message. This is using sqlite ver
On Tue, May 6, 2014 at 11:24 PM, Petite Abeille wrote:
> On May 6, 2014, at 11:17 PM, Richard Hipp wrote:
>
> > It is theoretically possible to keep track of which constraints are
> failing
> > so that the particular constraint can be identified in the error message.
> > But that woudl require mo
On May 6, 2014, at 11:17 PM, Richard Hipp wrote:
> It is theoretically possible to keep track of which constraints are failing
> so that the particular constraint can be identified in the error message.
> But that woudl require more memory and CPU cycles.
That would be resources well spent.
_
On Tue, May 6, 2014 at 5:12 PM, Peter Haworth wrote:
> It seems that foreign key errors on columns where the foreign key
> definition has a constraint name don't include the constraint name in the
> error message. This is using sqlite version 3.8.3.1.
>
> Is this under the control of a compile s
It seems that foreign key errors on columns where the foreign key
definition has a constraint name don't include the constraint name in the
error message. This is using sqlite version 3.8.3.1.
Is this under the control of a compile switch or PRAGMA or am I stuck with
the way it is?
Pete
14 matches
Mail list logo
