Add type checks for unsupported expressions in select and sort methods

Author: kosiew

python/datafusion/dataframe.py

@@ -400,7 +400,19 @@ def select(self, *exprs: Expr | str) -> DataFrame:
             df = df.select("a", col("b"), col("a").alias("alternate_a"))
 
         """
-        exprs_internal = _to_expr_list(exprs)
+        checked_exprs: list[Expr | str] = []
+        for expr in exprs:
+            if isinstance(expr, SortExpr):
+                checked_exprs.append(expr.expr())
+            elif isinstance(expr, (Expr, str)):
+                checked_exprs.append(expr)
+            else:
+                msg = (
+                    f"Expected Expr or column name, got {type(expr).__name__}. "
+                    "Use col() or lit() to construct expressions."
+                )
+                raise TypeError(msg)
+        exprs_internal = _to_expr_list(checked_exprs)
         return DataFrame(self.df.select(*exprs_internal))
 
     def drop(self, *columns: str) -> DataFrame:
@@ -540,9 +552,20 @@ def sort(self, *exprs: Expr | SortExpr | str) -> DataFrame:
         Returns:
             DataFrame after sorting.
         """
-        expr_seq = [e for e in exprs if not isinstance(e, SortExpr)]
+        expr_seq: list[Expr | str] = []
+        for e in exprs:
+            if isinstance(e, SortExpr):
+                continue
+            if isinstance(e, (Expr, str)):
+                expr_seq.append(e)
+            else:
+                msg = (
+                    f"Expected Expr or column name, got {type(e).__name__}. "
+                    "Use col() or lit() to construct expressions."
+                )
+                raise TypeError(msg)
         raw_exprs_iter = iter(_to_expr_list(expr_seq))
-        exprs_raw = []
+        exprs_raw: list[Any] = []
         for e in exprs:
             if isinstance(e, SortExpr):
                 exprs_raw.append(sort_or_default(e))

python/tests/test_dataframe.py

@@ -227,6 +227,13 @@ def test_select_mixed_expr_string(df):
     assert result.column(1) == pa.array([1, 2, 3])
 
 
+def test_select_unsupported(df):
+    with pytest.raises(
+        TypeError, match=r"Expected Expr or column name.*col\(\) or lit\(\)"
+    ):
+        df.select(1)
+
+
 def test_filter(df):
     df1 = df.filter(column("a") > literal(2)).select(
         column("a") + column("b"),
@@ -276,6 +283,13 @@ def test_sort_string_and_expression_equivalent(df):
     assert result_str == result_expr
 
 
+def test_sort_unsupported(df):
+    with pytest.raises(
+        TypeError, match=r"Expected Expr or column name.*col\(\) or lit\(\)"
+    ):
+        df.sort(1)
+
+
 def test_aggregate_string_and_expression_equivalent(df):
     from datafusion import col