Refactor sorting logic in DataFrame to handle string and Expr types for improved error handling and clarity

Author: kosiew

python/datafusion/dataframe.py

@@ -552,14 +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)]
-        raw_exprs_iter = iter(expr_list_to_raw_expr_list(expr_seq))
         exprs_raw = []
         for e in exprs:
             if isinstance(e, SortExpr):
                 exprs_raw.append(sort_or_default(e))
+            elif isinstance(e, str):
+                exprs_raw.append(sort_or_default(Expr.column(e)))
+            elif isinstance(e, Expr):
+                exprs_raw.append(sort_or_default(e))
             else:
-                exprs_raw.append(sort_or_default(Expr(next(raw_exprs_iter))))
+                error = (
+                    "Expected Expr or column name, found:"
+                    f" {type(e).__name__}. Use col() or lit() to construct expressions."
+                )
+                raise TypeError(error)
         return DataFrame(self.df.sort(*exprs_raw))
 
     def cast(self, mapping: dict[str, pa.DataType[Any]]) -> DataFrame: