BUG:Sanity check on merge parameters for correct exception #26824

doc/source/whatsnew/v0.25.0.rst

@@ -99,6 +99,7 @@ Other Enhancements
 - Error message for missing required imports now includes the original import error's text (:issue:`23868`)
 - :class:`DatetimeIndex` and :class:`TimedeltaIndex` now have a ``mean`` method (:issue:`24757`)
 - :meth:`DataFrame.describe` now formats integer percentiles without decimal point (:issue:`26660`)
+- :func:`pandas.merge` now raises ``ValueError`` when either of ``left_on`` or ``right_on`` is not provided (:issue:`26855`)
 
 .. _whatsnew_0250.api_breaking:
 

pandas/core/reshape/merge.py

@@ -1089,13 +1089,15 @@ def _validate_specification(self):
                     raise ValueError('len(left_on) must equal the number '
                                      'of levels in the index of "right"')
                 self.right_on = [None] * n
+            self.right_on = self.right_on or [None] * len(self.left_on)
         elif self.right_on is not None:
             n = len(self.right_on)
             if self.left_index:
                 if len(self.right_on) != self.left.index.nlevels:
                     raise ValueError('len(right_on) must equal the number '
                                      'of levels in the index of "left"')
                 self.left_on = [None] * n
+            self.left_on = self.left_on or [None] * len(self.right_on)
         if len(self.right_on) != len(self.left_on):
             raise ValueError("len(right_on) must equal len(left_on)")
 

pandas/tests/reshape/merge/test_merge.py

@@ -1763,3 +1763,20 @@ def test_merge_equal_cat_dtypes2():
 
     # Categorical is unordered, so don't check ordering.
     tm.assert_frame_equal(result, expected, check_categorical=False)
+
+
+@pytest.mark.parametrize('merge_type', ['left_on', 'right_on'])
+def test_missing_on_raises(merge_type):
+    # GH26824
+    df1 = DataFrame({
+        'A': [1, 2, 3, 4, 5, 6],
+        'B': ['P', 'Q', 'R', 'S', 'T', 'U']
+    })
+    df2 = DataFrame({
+        'A': [1, 2, 4, 5, 7, 8],
+        'C': ['L', 'M', 'N', 'O', 'P', 'Q']
+    })
+    msg = 'both left_on and right_on should be passed'
+    kwargs = {merge_type: 'A'}
+    with pytest.raises(ValueError, match=msg):
+        pd.merge(df1, df2, how='left', **kwargs)