DE-BB: leave out landscape elements, and map the zero-padded codes to HCAT

CHANGELOG.md

@@ -104,6 +104,7 @@ and this project adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.
 - CH: Removed the national `ch` converter, whose single licence could not cover the cantons' differing terms; the canton converters replace it, and a Swiss file is `fiboa merge` of their outputs.
 
 ### Fixed
+- DE-BB: Landscape elements (hedges, tree rows, copses, wetlands) are left out, and codes under 100, such as the eco-scheme areas, now map to HCAT; they come zero-padded in the source.
 - Added HCAT spelling fixes via `csv_supplements` for DE-BB, DE-NDS and EC-SI.
 - EC-FR, EC-NL-CROP: HCAT mappings for iris, Jerusalem artichoke and parsley now match the taxonomy.
 - EuroCrops converters whose source already carries HCAT (such as EC-FR and EC-SI) now apply their supplements too.

fiboa_cli/datasets/de_bb.py

@@ -37,4 +37,8 @@ class Converter(AdminConverterMixin, AddHCATMixin, FiboaBaseConverter):
         "code": "crop:code",
     }
     # NBF: areas in a field block the cadastre marks as not usable for agriculture (pylons, ditches, woods); no crop code
-    column_filters = {"tf_typ": lambda col: col != "NBF"}
+    # LE: landscape elements (hedges, tree rows, copses, wetlands), which are no fields; their
+    # codes 001-013 also name non-crop uses, so only the type tells them apart
+    column_filters = {"tf_typ": lambda col: ~col.isin(["NBF", "LE"])}
+    # codes under 100 come zero-padded (088), the code lists write them without (88)
+    column_migrations = {"code": lambda col: col.str.replace(r"^0+(?=\d)", "", regex=True)}

tests/test_convert.py

@@ -363,8 +363,13 @@ def _input_files(converter, *names):
     "ec_be_vlg": ("1", "2"),
     # small woody features and other areas that are no field
     "de_thuenen": ("3001", "3004"),
+    # landscape elements (tf_typ LE); in the fixture these codes name nothing else
+    "de_bb": ("1", "3", "4", "11", "13", "001", "003", "004", "011", "013"),
 }
 
+# Converters whose fixture maps every row to HCAT
+fully_mapped = {"de_bb"}
+
 # The smallest field (m²) a converter keeps, where the fixture holds a smaller sliver
 minimum_area = {
     "br_conab": 100,
@@ -444,6 +449,10 @@ def load_mapping(csv_file=None, url=None):
     dropped = [c for c in filtered_codes.get(converter, ()) if c in set(df["crop:code"])]
     assert not dropped, f"{converter} kept rows it should filter out: crop:code {dropped}"
 
+    if converter in fully_mapped:
+        unmapped = sorted(set(df.loc[df["hcat:code"].isna(), "crop:code"]))
+        assert not unmapped, f"{converter}: crop:code {unmapped} has no HCAT code"
+
     if "metrics:area" in df.columns and converter not in ("de_bb", "ec_de_bb"):
         # Check for accidental hectare conversion; fields should be more than 10 square meters
         assert (df["metrics:area"] > 10).all()