qgepqwat2ili: exported labels' TID are no standardoid

[cymed]

Describe the bug When exporting a Model including labels, the labels have a purely numerical TID .

To Reproduce Exact steps to reproduce the behavior:

1. Export an xtf with labels through the GUI
2. Check a feature's TID in class 'xx_Text'
3. TID is numeric

Expected behavior TID is a valid standardoid as required by SIA405_Base. My take on this would be:

1. Separate the label export from the interlis export 1.1. Add a DB- side view tww_od.vw_text for import and export 1.2. Add a UI to import labels into tww_od.txt_text and the corresponding specialised table (i.e. tww_od.reach_text) 1.3 Potential problem: We are currently limited to 1 million OID per class, see https://github.com/QGEP/QGEP/discussions/764. When we truncate and re-import the labels every time we want to make an export, this might be reached quite fast
2. Altering the export scripts to use the label positions in the database instead of the ones in the QGIS project. Importing the labels into

Screenshots / data

Desktop (please complete the following information):

• QGEP plugin version: 1.6.1
• QGEP datamodel version: 1.6.0 with necessary PR for VSA DSS 2015 Export (https://github.com/QGEP/qgepqwat2ili/tree/2023-03_interlis_additional_models_new)
• QGEP project file version: 1.6.0
• QGIS Version: 3.28.11
• OS: Windows

Additional context Add any other context about the problem here.

[sjib]

@cymed

• Label export was not intended to do a data exchange but export only for geoportals that cannot place labels in a optimized position.
• Do you get validation errors in INTERLIS with numeric TID?
• If yes, with which datamodel(s)?

[cymed]

I get no validation errors from ilivalidator, but the canton of Aargau requires all tids to be standardoids according to an extended validation check (relevant for AG64/96 gateway that is in the planning)

[sjib]

But label_position are a "waste" product that will change with a next data delivery and cannot be maintained - so this makes no sense to me to set a stable OID there.

[cymed]

I agree conceptually, but as long as some cantonal offices require stable oid, the engineering companies must deliver them.

[cymed]

An alternative would be to add the labels to mapping.py and decide on the population mechanism (from db or from geojson) by a flag

[sjib]

@cymed Do we still have to solve this in qgep? Else please close and add issue to tww

[cymed]

Is solved already in tww. Closing here