enforce limitation to polygon geometries cf. #5 #6

vector2dggs/h3.py

@@ -102,7 +102,7 @@ def polyfill_star(args) -> None:
 def _parent_partitioning(
     input_dir: Path,
     output_dir: Path,
-    resolution,
+    resolution: int,
     parent_res: Union[None, int],
     **kwargs,
 ) -> None:
@@ -128,6 +128,16 @@ def _parent_partitioning(
 
     return
 
+def drop_condition(df : pd.DataFrame, drop_index : pd.Index, log_statement : str, warning_threshold : float =0.01):
+    LOGGER.info(log_statement)
+    _before = len(df)
+    df = df.drop(drop_index)
+    _after = len(df)
+    _diff = _before - _after
+    if _diff:
+        log_method = LOGGER.info if (_diff/float(_before)) < warning_threshold else LOGGER.warn
+        log_method(f"Dropped {_diff} rows ({_diff/float(_before)*100:.2f}%)")
+    return df
 
 def _index(
     input_file: Union[Path, str],
@@ -193,12 +203,19 @@ def _index(
         .explode(index_parts=False)  # Explode from GeometryCollection
         .explode(index_parts=False)  # Explode multipolygons to polygons
     ).reset_index()
-    LOGGER.info("Dropping empty or null geometries")
-    df = (
-        df.drop(df[(df.geometry.is_empty|df.geometry.isna())].index)
-        .reset_index()
-    )
 
+    drop_conditions = [
+        {
+            'index': lambda frame: frame[(frame.geometry.is_empty|frame.geometry.isna())],
+            'message': "Dropping empty or null geometries"
+        }, {
+            'index': lambda frame: frame[(frame.geometry.geom_type!="Polygon")], # NB currently points and lines are lost; in principle, these could be indexed
+            'message': "Dropping non-polygonal geometries"
+        }
+    ]
+    for condition in drop_conditions:
+        df = drop_condition(df, condition['index'](df).index, condition['message'])
+
     ddf = dgpd.from_geopandas(df, chunksize=max(1, chunksize), sort=True)
 
     LOGGER.info("Spatially sorting and partitioning (%s)", spatial_sorting)