Hi,
Yesterday I added some signals to QgsVectorLayer which are emitted in
calls to beginEditCommand, endEditCommand and destroyEditCommand. This
makes it possible for consumers of edit signals (attributeValueChanged
etc.) to wait for a batch operation to finish and update once at the
end. Example: In the case of the attribute table and an update to all
columns from the field calculator it results in an enormous speed boost,
if it is updated only once at the end of an operation and not for every
single feature.
What I am doing now is to cache the "bounds" (min/max column/row) of
changed attributes and when the next endEditCommand is made update
everything within.
Now the question is: Can I rely on endEditCommand being called for every
operation? Or do I have to check for out-of-edit-command changes and
update immediately in this case?
Regards,
Matthias
_______________________________________________
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer
