[Qgis-user] Split features using Geopackage layer?

[Nyall Dawson]

On 19 October 2016 at 22:55, Neumann, Andreas <a.neumann at carto.net> wrote:
> As far as I know, Nyall Dawson is currently working on fixing this, as part
> of his work on unique value constraints (provider provided or qgis provided
> where the provider doesn't support it). But not sure about the details.
> Certainly, PostgreSQL and Geopackage are top priorities for that. And his
> work targets QGIS 3.x, not for 2.x

Yes, I'm currently working on this as part of a larger work in
auto-exposing provider side constraints (not null/unique) to QGIS's
form validation.

WIP is here: https://github.com/nyalldawson/QGIS/commits/constraints,
but so far it only handles the detection + translation of constraints
from provider -> QGIS. The form based/tool changes are still to come.

Nyall


>
> Greetings,
>
> Andreas
>
>
>
> On 2016-10-19 14:43, Even Rouault wrote:
>
> Le mercredi 19 octobre 2016 14:14:40, Chris Buckmaster a écrit :
>
> Hi
>
> I have a geopackage polygon layer and I am trying to split a simple feature
> within it.
>
> When I try to do it, I get an error message 'failed to insert, UNIQUE
> constraint failed'.
>
> Is this a known issue with splitting features using the geopackage format,
> it could it be how the layer is set up?
>
>
> No, it is due to the fact that when splitting the feature, the newly created
> feature receives the same 'fid' value (the first column in the attribute
> table)
> as the other part, and that when creating the feature, the fid, when set, is
> honoured. Hence given the duplication of fid, the failure. Before saving the
> edits, you can manually unset the fid.
>
> Although something should probably be done either in the splitting tool to
> unset the fid column of the new feature. Or perhaps the provider could be
> made
> more tolerant in the case of duplicated FID to assign a new one.
>
> This could happen for other formats handled by OGR like spatialite (if not
> going throug the QGIS spatialite provider), FileGDB, etc...
>
> Would probably deserve a ticket in the bug tracker.
>
> Even
>
>
>
>
>
> _______________________________________________
> Qgis-user mailing list
> Qgis-user at lists.osgeo.org
> List info: http://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: http://lists.osgeo.org/mailman/listinfo/qgis-user