Fix bug when loading few columns of a dataset with many primary indic…

…es (JDASoftwareGroup#446)

* Only increase pos if necessary.

* Changelog.

* Empty line in changelog.

* Add test.

* Don't link to private function.

CHANGES.rst

@@ -2,6 +2,13 @@
 Changelog
 =========
 
+
+Kartothek 4.0.2 (2021-04-xx)
+============================
+
+* Fix a bug in ``MetaPartition._reconstruct_index_columns`` that would raise an ``IndexError`` when loading few columns of a dataset with many primary indices.
+
+
 Kartothek 4.0.1 (2021-04-13)
 ============================
 

kartothek/io_components/metapartition.py

@@ -767,7 +767,8 @@ def _reconstruct_index_columns(
             # indexer call is slow, so only do that if really necessary
             df = df.reindex(columns=cleaned_original_columns, copy=False)
 
-        for pos, (primary_key, value) in enumerate(key_indices):
+        pos = 0
+        for primary_key, value in key_indices:
             # If there are predicates, don't reconstruct the index if it wasn't requested
             if columns is not None and primary_key not in columns:
                 continue
@@ -801,6 +802,7 @@ def _reconstruct_index_columns(
                 if convert_to_date:
                     value = pd.Timestamp(value).to_pydatetime().date()
             df.insert(pos, primary_key, value)
+            pos += 1
 
         return df
 

tests/io_components/test_metapartition.py

@@ -27,7 +27,7 @@ def test_store_single_dataframe_as_partition(store, metadata_version):
     mp = MetaPartition(label="test_label", data=df, metadata_version=metadata_version)
 
     meta_partition = mp.store_dataframes(
-        store=store, df_serializer=ParquetSerializer(), dataset_uuid="dataset_uuid",
+        store=store, df_serializer=ParquetSerializer(), dataset_uuid="dataset_uuid"
     )
 
     assert meta_partition.data is None
@@ -58,7 +58,7 @@ def test_load_dataframe_logical_conjunction(store, metadata_version):
         logical_conjunction=[("P", ">", 4)],
     )
     meta_partition = mp.store_dataframes(
-        store=store, df_serializer=None, dataset_uuid="dataset_uuid",
+        store=store, df_serializer=None, dataset_uuid="dataset_uuid"
     )
     predicates = None
     loaded_mp = meta_partition.load_dataframes(store=store, predicates=predicates)
@@ -1333,6 +1333,20 @@ def test_get_parquet_metadata_row_group_size(store):
     pd.testing.assert_frame_equal(actual, expected)
 
 
+def test__reconstruct_index_columns():
+    df = pd.DataFrame({"x": [0], "a": [-1], "b": [-2], "c": [-3]})
+    mp = MetaPartition(label="test_label", data=df)
+    df_with_index_columns = mp._reconstruct_index_columns(
+        df=df[["x"]],
+        key_indices=[("a", 1), ("b", 2), ("c", 3)],
+        columns=["x", "c"],
+        categories=None,
+        date_as_object=False,
+    )
+    # Index columns first
+    pdt.assert_frame_equal(df_with_index_columns, pd.DataFrame({"c": [3], "x": [0]}))
+
+
 def test_partition_on_keeps_table_name():
     mp = MetaPartition(
         label="label_1",