Two version of a dataset

[ankitkariryaa]

I have two versions of a dataset. Let's say d_v1.gpkg and d_v2.gpkg. Now I created a kart repo with kart init dtt --import GPKG:d_v1.gpkg.

Now I would like to add d_v2.gpkg to the same source. How do I do that with kart?

[rcoup]

You can use kart import --replace-existing — this will load in d_v2.gpkg as a new revision of the same dataset, but identical records from the existing data get (effectively) matched so kart log/kart diff etc will only show the differences.

Note that for this to work properly, you really want a consistent primary key across your import sources. Kart tries to auto-detect this (eg: GeoPackage mandates you have a primary key), but you can also manually specify it via --primary-key.

[ankitkariryaa]

Okay, that's great, thanks!