NO-SNOW: fix release test failure (#4191)

Author: sfc-gh-aling

tests/ast/test_ast_driver.py

@@ -250,9 +250,15 @@ def test_ast(session, tables, test_case, resources_path):
             ClearTempTables(actual_message)
             ClearTempTables(expected_message)
 
-            # If this is not python 3.11+, then for the purposes of the expectation tests we will ignore the line_no
-            # information since it can be different based on various python bug fixes.
-            if sys.version_info.minor <= 10:
+            # Ignore src line/column positions when they are not stable across interpreters:
+            # - Python < 3.11 had inconsistent frame line numbers (see clear_line_no_in_request).
+            # - Python 3.14+ can disagree with the golden files (typically produced on 3.11/3.12) for the same
+            #   user code, e.g. due to changes in how frames map to the temp test wrapper (similar to df_copy
+            #   being skipped on 3.12 in load_test_cases).
+            # TODO SNOW-3414082: Golden-file improvements so we can assert src positions across interpreters without clearing
+            # them here—e.g. version-specific expected AST files, normalizing only wrapper-related offsets, or
+            # recording positions relative to trimmed user code.
+            if sys.version_info < (3, 11) or sys.version_info >= (3, 14):
                 clear_line_no_in_request(actual_message)
                 clear_line_no_in_request(expected_message)
 

tests/integ/modin/test_faster_pandas.py

@@ -83,6 +83,10 @@ def test_read_filter_join(session):
         assert_frame_equal(snow_result, native_result)
 
 
+@pytest.mark.xfail(
+    reason="SNOW-3415226: flaky test_read_filter_join_on_index",
+    strict=False,
+)
 @sql_count_checker(query_count=6, join_count=2)
 def test_read_filter_join_on_index(session):
     with session_parameter_override(

tests/integ/scala/test_dataframe_reader_suite.py

@@ -47,6 +47,7 @@
     TimeType,
     VariantType,
 )
+from snowflake.snowpark._internal.server_connection import MAX_STRING_SIZE
 from tests.utils import (
     IS_IN_STORED_PROC,
     IS_IN_STORED_PROC_LOCALFS,
@@ -387,7 +388,9 @@ def test_read_csv_with_user_schema_try_cast(session, mode):
     try_cast_schema = StructType(
         [
             StructField("A", LongType()),
-            StructField("B", StringType()),
+            StructField(
+                "B", StringType(length=MAX_STRING_SIZE)
+            ),  # Standardizing string size to 128MB for regression test env parity.
             StructField("C", DoubleType()),
         ]
     )

tests/integ/scala/test_dataframe_writer_suite.py

@@ -247,15 +247,17 @@ def test_iceberg(session, local_testing_mode):
     )
     try:
         ddl = session._run_query(f"select get_ddl('table', '{table_name}')")
-        assert (
-            ddl[0][0] == f"create or replace ICEBERG TABLE {table_name} (\n\t"
-            f"A STRING,\n\tB LONG,\n\tTS TIMESTAMP_NTZ(9)\n)\n "
+        # TODO: AWS default may emit TIMESTAMP_NTZ(6) instead of (9).
+        assert ddl[0][0] in {
+            f"create or replace ICEBERG TABLE {table_name} (\n\t"
+            f"A STRING,\n\tB LONG,\n\tTS TIMESTAMP_NTZ({prec})\n)\n "
             f"PARTITION BY (A, BUCKET(5, B), TRUNCATE(3, A), DAY(TS))\n "
             f"EXTERNAL_VOLUME = 'PYTHON_CONNECTOR_ICEBERG_EXVOL'\n "
             f"ICEBERG_VERSION = 3\n "
             f"CATALOG = 'SNOWFLAKE'\n "
             f"BASE_LOCATION = 'snowpark_python_tests/';"
-        )
+            for prec in (9, 6)
+        }
 
         params = session.sql(f"show parameters for table {table_name}").collect()
         target_file_size_params = [

tests/integ/test_df_aggregate.py

@@ -962,6 +962,10 @@ def test_group_by_agg_sort_filter_sanity(session):
         GROUP BY "DEPT"
         """
 
+        integer_literal_postfix = (
+            "" if session.eliminate_numeric_sql_value_cast_enabled else " :: INT"
+        )
+
         def check_agg_sql(df, expected_sql):
             assert Utils.normalize_sql(df.queries["queries"][0]) == Utils.normalize_sql(
                 expected_sql
@@ -980,7 +984,7 @@ def check_agg_sql(df, expected_sql):
             result1,
             f"""
             {agg_query_base}
-            HAVING ("HEADCOUNT" > 1)
+            HAVING ("HEADCOUNT" > 1{integer_literal_postfix})
             ORDER BY "AVG_SALARY" DESC NULLS LAST
             """,
         )
@@ -992,7 +996,7 @@ def check_agg_sql(df, expected_sql):
             result2,
             f"""
             {agg_query_base}
-            HAVING ("HEADCOUNT" > 1)
+            HAVING ("HEADCOUNT" > 1{integer_literal_postfix})
             ORDER BY "AVG_SALARY" DESC NULLS LAST
             """,
         )
@@ -1014,7 +1018,7 @@ def check_agg_sql(df, expected_sql):
             result3,
             f"""
             {agg_query_base}
-            HAVING ("HEADCOUNT" > 1)
+            HAVING ("HEADCOUNT" > 1{integer_literal_postfix})
             ORDER BY "AVG_SALARY" DESC NULLS LAST
             LIMIT 2 OFFSET 0
             """,
@@ -1038,7 +1042,7 @@ def check_agg_sql(df, expected_sql):
                 {agg_query_base}
                 ORDER BY "AVG_SALARY" DESC NULLS LAST
             )
-            WHERE ("HEADCOUNT" > 1)
+            WHERE ("HEADCOUNT" > 1{integer_literal_postfix})
             """
                 if session.sql_simplifier_enabled
                 else f"""
@@ -1049,7 +1053,7 @@ def check_agg_sql(df, expected_sql):
                     ORDER BY "AVG_SALARY" DESC NULLS LAST
                 )
             )
-            WHERE ("HEADCOUNT" > 1)
+            WHERE ("HEADCOUNT" > 1{integer_literal_postfix})
             """
             ),
         )
@@ -1074,7 +1078,7 @@ def check_agg_sql(df, expected_sql):
             result5,
             f"""
             {agg_query_base}
-            HAVING ("HEADCOUNT" > 1)
+            HAVING ("HEADCOUNT" > 1{integer_literal_postfix})
             ORDER BY "HEADCOUNT" ASC NULLS FIRST,
                 "AVG_SALARY" DESC NULLS LAST,
                 "AVG_SALARY" ASC NULLS FIRST
@@ -1098,7 +1102,7 @@ def check_agg_sql(df, expected_sql):
             result6,
             f"""
             {agg_query_base}
-            HAVING (("HEADCOUNT" > 1) AND ("AVG_SALARY" > 50000))
+            HAVING (("HEADCOUNT" > 1{integer_literal_postfix}) AND ("AVG_SALARY" > 50000{integer_literal_postfix}))
             ORDER BY "AVG_SALARY" DESC NULLS LAST
             """,
         )

tests/integ/test_lineage.py

@@ -61,6 +61,7 @@ def convert_and_clean(json_str):
 
 
 @pytest.mark.skipif(not is_pandas_available, reason="pandas is required")
+@pytest.mark.skip(reason="SNOW-3413244 lineage test is flaky")
 def test_lineage_trace(session):
     """
     Tests lineage.trace API on multiple cases.