temporal dbif for current mapset only

[metzm]

Sometimes only the current mapset must be used by TGIS, particularly if the TGIS db is going to be modified. TGIS dbs in other mapsets can/should not be modified, only queried.

This PR adds functionality to initialize a SQLDatabaseInterfaceConnection with only the current mapset.

[wenzeslaus]

This looks reasonable. I have some documentation and stylistic notes.

[wenzeslaus]

As far as the code goes, this is good to go.

[metzm]

A related test for TGIS mapset access is failing. This needs to be fixed first before merging.

[metzm]

A related test for TGIS mapset access is failing. This needs to be fixed first before merging.

This will require a bit more effort because GRASS TGIS still expects at various different places that a raster/raster3d/vector map can only be registered in a TGIS db in the same mapset where this map is located. This prevents registering maps from a different mapset in a space-time dataset in the current mapset and subsequent modifications of this space-time dataset in the current mapset.

A solution would be to not only store information about the mapset of a map, but also information about the mapset of the space-time dataset that is currently used and where this map is registered in the internal structures of the temporal framework.

The historical background is that a map could only be registered in a space-time dataset if both are in the current mapset.

[wenzeslaus]

At this point, this seems like something for 8.4.0.

[neteler]

@metzm shall I bump this PR to 8.4.0?

[landam]

@metzm Please consider merging this PR.

[echoix]

I tried to solve conflicts, but I couldn't make sense of the changes in regards to changing self.mapset to self.data_mapset, that didn't seem used correctly. So I aborted it.

[ninsbl]

There have been a couple of recent changes that overlap with this PR:
#7707
#7672
#7671
And there is pending: #7670 which looks at this from a different perspective...
Will look into this as part of the Community meeting.