fix: Keeping no columns with Table.keep_only_columns results in an em…

…pty Table with a row count above 0 (#386)

Closes #318 

Fixes Bug:

If you use the method Table.keep_only_columns with an empty list on a
table that contains at least one row, you get an empty table with a row
count above 0.

### Summary of Changes

<!-- Please provide a summary of changes in this pull request, ensuring
all changes are explained. -->

---------

Co-authored-by: patrikguempel <[email protected]>
Co-authored-by: megalinter-bot <[email protected]>
Co-authored-by: Simon Breuer <[email protected]>

src/safeds/data/tabular/containers/_table.py

@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import copy
 import functools
 import io
 import warnings
@@ -440,8 +441,8 @@ def __eq__(self, other: Any) -> bool:
             return NotImplemented
         if self is other:
             return True
-        if self.number_of_rows == 0 and other.number_of_rows == 0:
-            return self.column_names == other.column_names
+        if self.number_of_columns == 0 and other.number_of_columns == 0:
+            return True
         table1 = self.sort_columns()
         table2 = other.sort_columns()
         if table1.number_of_rows == 0 and table2.number_of_rows == 0:
@@ -1093,9 +1094,9 @@ def keep_only_columns(self, column_names: list[str]) -> Table:
         if len(invalid_columns) != 0:
             raise UnknownColumnNameError(invalid_columns)
 
-        transformed_data = self._data[column_names]
-        transformed_data.columns = column_names
-        return Table._from_pandas_dataframe(transformed_data)
+        clone = copy.deepcopy(self)
+        clone = clone.remove_columns(list(set(self.column_names) - set(column_names)))
+        return clone
 
     def remove_columns(self, column_names: list[str]) -> Table:
         """
@@ -1138,6 +1139,10 @@ def remove_columns(self, column_names: list[str]) -> Table:
 
         transformed_data = self._data.drop(labels=column_names, axis="columns")
         transformed_data.columns = [name for name in self._schema.column_names if name not in column_names]
+
+        if len(transformed_data.columns) == 0:
+            return Table()
+
         return Table._from_pandas_dataframe(transformed_data)
 
     def remove_columns_with_missing_values(self) -> Table:

tests/safeds/data/tabular/containers/_table/test_eq.py

@@ -8,14 +8,16 @@
     ("table1", "table2", "expected"),
     [
         (Table(), Table(), True),
+        (Table({"a": [], "b": []}), Table({"a": [], "b": []}), True),
         (Table({"col1": [1]}), Table({"col1": [1]}), True),
         (Table({"col1": [1]}), Table({"col2": [1]}), False),
         (Table({"col1": [1, 2, 3]}), Table({"col1": [1, 1, 3]}), False),
         (Table({"col1": [1, 2, 3]}), Table({"col1": ["1", "2", "3"]}), False),
     ],
     ids=[
-        "empty Table",
-        "equal Tables",
+        "empty table",
+        "rowless table",
+        "equal tables",
         "different column names",
         "different values",
         "different types",

tests/safeds/data/tabular/containers/_table/test_keep_only_columns.py

@@ -9,7 +9,7 @@
         (
             Table({"A": [1], "B": [2]}),
             [],
-            Table({}),
+            Table(),
         ),
         (
             Table({"A": [1], "B": [2]}),
@@ -44,6 +44,8 @@ def test_should_keep_only_listed_columns(table: Table, column_names: list[str],
     transformed_table = table.keep_only_columns(column_names)
     assert transformed_table.schema == expected.schema
     assert transformed_table == expected
+    if len(column_names) == 0:
+        assert expected.number_of_rows == 0
 
 
 @pytest.mark.parametrize("table", [Table({"A": [1], "B": [2]}), Table()], ids=["table", "empty"])

tests/safeds/data/tabular/containers/_table/test_remove_columns.py

@@ -4,19 +4,20 @@
 
 
 @pytest.mark.parametrize(
-    ("table1", "expected", "columns"),
+    ("table", "expected", "columns"),
     [
         (Table({"col1": [1, 2, 1], "col2": ["a", "b", "c"]}), Table({"col1": [1, 2, 1]}), ["col2"]),
-        (Table({"col1": [1, 2, 1], "col2": [1, 2, 4]}), Table({}), ["col1", "col2"]),
+        (Table({"col1": [1, 2, 1], "col2": [1, 2, 4]}), Table(), ["col1", "col2"]),
         (Table({"col1": [1, 2, 1], "col2": [1, 2, 4]}), Table({"col1": [1, 2, 1], "col2": [1, 2, 4]}), []),
         (Table(), Table(), []),
     ],
     ids=["one column", "multiple columns", "no columns", "empty"],
 )
-def test_should_remove_table_columns(table1: Table, expected: Table, columns: list[str]) -> None:
-    table1 = table1.remove_columns(columns)
-    assert table1.schema == expected.schema
-    assert table1 == expected
+def test_should_remove_table_columns(table: Table, expected: Table, columns: list[str]) -> None:
+    table = table.remove_columns(columns)
+    assert table.schema == expected.schema
+    assert table == expected
+    assert table.number_of_rows == expected.number_of_rows
 
 
 @pytest.mark.parametrize("table", [Table({"A": [1], "B": [2]}), Table()], ids=["normal", "empty"])