minor comments

Author: sfc-gh-joshi

src/snowflake/snowpark/_internal/analyzer/select_statement.py

@@ -2250,6 +2250,7 @@ def derive_column_states_from_subquery(
         if from_c_state and from_c_state.change_state != ColumnChangeState.DROPPED:
             # review later. should use parse_column_name
             # SNOW-2895675: Always treat Aliases as "changed", even if it is an identity.
+            # The fact this check is needed may be a bug in column state analysis, and we should revisit it later.
             if c_name != result_name or isinstance(c, Alias):
                 column_states[quoted_c_name] = ColumnState(
                     quoted_c_name,

src/snowflake/snowpark/dataframe.py

@@ -346,7 +346,8 @@ def _disambiguate(
         # Optimization: No column name conflicts, so we can skip aliasing and the select() wrapping.
         # But we still need to populate expr_to_alias for column lineage tracking,
         # so that df["column_name"] can resolve correctly after the join.
-        # This is identified by the test case test_name_alias_on_multiple_join.
+        # This is identified by the test case
+        # tests/integ/scala/test_dataframe_join_suite.py::test_name_alias_on_multiple_join.
         _populate_expr_to_alias(lhs)
         _populate_expr_to_alias(rhs)
         return lhs, rhs