BUG: MergeError not raised when both right_on and right_index are specified (#63248)

Author: lexzlei

doc/source/whatsnew/v3.0.0.rst

@@ -1346,6 +1346,7 @@ Reshaping
 - Bug in :meth:`DataFrame.unstack` producing incorrect results when manipulating empty :class:`DataFrame` with an :class:`ExtentionDtype` (:issue:`59123`)
 - Bug in :meth:`concat` where concatenating DataFrame and Series with ``ignore_index = True`` drops the series name (:issue:`60723`, :issue:`56257`)
 - Bug in :func:`melt` where calling with duplicate column names in ``id_vars`` raised a misleading ``AttributeError`` (:issue:`61475`)
+- Bug in :meth:`DataFrame.merge` where specifying both ``right_on`` and ``right_index`` did not raise a ``MergeError`` if ``left_on`` is also specified. Now raises a ``MergeError`` in such cases. (:issue:`63242`)
 - Bug in :meth:`DataFrame.merge` where user-provided suffixes could result in duplicate column names if the resulting names matched existing columns. Now raises a :class:`MergeError` in such cases. (:issue:`61402`)
 - Bug in :meth:`DataFrame.merge` with :class:`CategoricalDtype` columns incorrectly raising ``RecursionError`` (:issue:`56376`)
 - Bug in :meth:`DataFrame.merge` with a ``float32`` index incorrectly casting the index to ``float64`` (:issue:`41626`)

pandas/core/reshape/merge.py

@@ -1928,6 +1928,10 @@ def _validate_left_right_on(self, left_on, right_on):
                 )
             if not self.right_index and right_on is None:
                 raise MergeError('Must pass "right_on" OR "right_index".')
+            if self.right_index and right_on is not None:
+                raise MergeError(
+                    'Can only pass argument "right_on" OR "right_index" not both.'
+                )
             n = len(left_on)
             if self.right_index:
                 if len(left_on) != self.right.index.nlevels:

pandas/tests/reshape/merge/test_merge.py

@@ -3149,3 +3149,11 @@ def test_merge_pyarrow_datetime_duplicates():
     )
     expected = expected.convert_dtypes(dtype_backend="pyarrow")
     tm.assert_frame_equal(result, expected)
+
+
+def test_merge_right_on_and_right_index():
+    df1 = DataFrame({"col": [1, 2, 3]})
+    df2 = DataFrame({"col": [2, 3, 4]})
+
+    with pytest.raises(pd.errors.MergeError):
+        df1.merge(df2, left_on="col", right_on="col", right_index=True)