📋 Issue: data.world integration

[profsergiocosta]

Title:
[ENH] Re-add data.world integration in dissmodel.io

Labels:
enhancement data-world io sparql

Objective:
Restore the ability to load geospatial data directly from data.world via SPARQL queries, returning a GeoDataFrame ready for use in DisSModel.

Previous functionality (reference):

def dw_query(dataset, query, geo_column='wkt', query_type='sparql'):
    results = dw.query(dataset, query, query_type)
    gdf = gpd.GeoDataFrame(
        results.dataframe,
        geometry=gpd.GeoSeries.from_wkt(results.dataframe[geo_column])
    )
    return gdf

Initial ideas:

• Create dissmodel/io/dataworld.py with query functions
• Support for SPARQL and SQL (if applicable)
• Automatic WKT/WKB to geometry conversion
• Integration with vector_to_raster_backend for direct flow: data.world → GeoDataFrame → RasterBackend
• Credentials via environment variables (.env)

Next steps:

• Review current datadotworld API (changes since last version)
• Decide whether to keep as optional dependency
• Decide whether to expose only raw query or also specific helpers (e.g., load_dbcells())
• Implement and test with a real dataset

Notes:
Detailed implementation to be defined after investigation. Consider credential security and connection error handling.