Add node_id column to Node table

[Hofer-Julian]

Breaking ⚠️

• Document changes
• Make sure that existing scripts still work
• Adapt tests
• Add update script for
  • Adding node_id to Node table if it doesn't exist already
  • Renaming type to node_type in Node table

Upgrade information

import geopandas as gpd

path = "path/to/database.gpkg"

# Load the GeoPackage as a GeoDataFrame
gdf = gpd.read_file(path, layer="Node", fid_as_index=True)

# Rename the column "type" to "node_type"
gdf.rename(columns={"type": "node_type"}, inplace=True)

# Add a new column "node_id" with the values of the index
gdf["node_id"] = gdf.index

# Save it back to a GeoPackage
gdf.to_file(path, layer="Node", driver="GPKG")

[visr]

This is good to merge from my side, if you can still add upgrade information with a script, similar to #992.

[Hofer-Julian]

This is good to merge from my side, if you can still add upgrade information with a script, similar to #992.

You want me to add a script which purpose is to adapt an existing Ribasim model so that it can be read by an up-to-date version of Ribasim Python?

[visr]

Yeah, using Ribasim Python if possible or sqlite3 if not. I know it is a bit of a hassle, but otherwise it becomes impossible to debug user models.

[Hofer-Julian]

Yeah, using Ribasim Python if possible or sqlite3 if not. I know it is a bit of a hassle, but otherwise it becomes impossible to debug user models.

I can take a look, but would be curious about the motivation.
I assumed you can always just take an older Ribasim Python version.

[visr]

You can take an older Ribasim Python to load the model, but to debug possible issues we need to be able to run the core on the main branch.

Ideally it is always fully reproducible and we can just regenerate models with scripts, but that is often not possible. Only for our test models it is that simple.

I think at some point we should start making an upgrade script that automatically updates models to the current version. For that to work we should start adding ribasim_version to the TOML file as well.