geointent 0.1.1

Natural language to spatial geometry — compiles NL to PostGIS, GeoPandas, GeoJSON, DuckDB

geointent

Natural language → spatial geometry: turn plain-language location questions into PostGIS SQL, GeoPandas / Shapely code, GeoJSON-style filters, or DuckDB spatial SQL.

Install

pip install geointent

import geointent
from geointent import Dialect

engine = geointent.Engine(
    llm="claude",
    context=geointent.SpatialContext(
        schema=geointent.Schema.from_dict({...}),
        domain="utility_network",
        units="feet",
        srid=2965,
    ),
)
result = engine.translate("manholes within 50 feet of a gas line")
print(result.query)

• PyPI: geointent
• Source: github.com/FultonGeo/geointent

Optional extras: pip install geointent[postgis], [duckdb], [geopandas].

Three common use cases

1. Utility / asset network (PostGIS) — Projected CRS (e.g. state plane), mains and structures: Dialect.POSTGIS (default), parameterized SQL, optional Engine.execute with your SQLAlchemy connection.
2. Parcels & zones — Dialect.GEOJSON or Dialect.GEOPANDAS for polygon relationships and buffers.
3. Overture / GeoParquet (DuckDB) — Dialect.DUCKDB for bbox-safe SQL against remote Parquet.

Runnable demos (mock LLM, no API keys): python examples/utility_network.py, examples/overture_duckdb.py, examples/parcel_analysis.py.

Development (venv)

python -m venv .venv
.\.venv\Scripts\Activate.ps1
pip install -e ".[dev,postgis,geopandas]"
python -m pytest tests/ -m "not integration and not live and not realdata"

On macOS/Linux: source .venv/bin/activate.

See TESTING.md for PostGIS Docker setup and pytest tiers.