Disable JsonTuple by default (#10420)

Signed-off-by: Robert (Bobby) Evans <bobby@apache.org>

docs/additional-functionality/advanced_configs.md

@@ -268,7 +268,7 @@ Name | SQL Function(s) | Description | Default Value | Notes
 <a name="sql.expression.IsNotNull"></a>spark.rapids.sql.expression.IsNotNull|`isnotnull`|Checks if a value is not null|true|None|
 <a name="sql.expression.IsNull"></a>spark.rapids.sql.expression.IsNull|`isnull`|Checks if a value is null|true|None|
 <a name="sql.expression.JsonToStructs"></a>spark.rapids.sql.expression.JsonToStructs|`from_json`|Returns a struct value with the given `jsonStr` and `schema`|false|This is disabled by default because it is currently in beta and undergoes continuous enhancements. Please consult the [compatibility documentation](../compatibility.md#json-supporting-types) to determine whether you can enable this configuration for your use case|
-<a name="sql.expression.JsonTuple"></a>spark.rapids.sql.expression.JsonTuple|`json_tuple`|Returns a tuple like the function get_json_object, but it takes multiple names. All the input parameters and output column types are string.|true|None|
+<a name="sql.expression.JsonTuple"></a>spark.rapids.sql.expression.JsonTuple|`json_tuple`|Returns a tuple like the function get_json_object, but it takes multiple names. All the input parameters and output column types are string.|false|This is disabled by default because JsonTuple on the GPU does not support all of the normalization that the CPU supports.|
 <a name="sql.expression.KnownFloatingPointNormalized"></a>spark.rapids.sql.expression.KnownFloatingPointNormalized| |Tag to prevent redundant normalization|true|None|
 <a name="sql.expression.KnownNotNull"></a>spark.rapids.sql.expression.KnownNotNull| |Tag an expression as known to not be null|true|None|
 <a name="sql.expression.Lag"></a>spark.rapids.sql.expression.Lag|`lag`|Window function that returns N entries behind this one|true|None|

docs/supported_ops.md

@@ -8222,7 +8222,7 @@ are limited.
 <td rowSpan="3">JsonTuple</td>
 <td rowSpan="3">`json_tuple`</td>
 <td rowSpan="3">Returns a tuple like the function get_json_object, but it takes multiple names. All the input parameters and output column types are string.</td>
-<td rowSpan="3">None</td>
+<td rowSpan="3">This is disabled by default because JsonTuple on the GPU does not support all of the normalization that the CPU supports.</td>
 <td rowSpan="3">project</td>
 <td>json</td>
 <td> </td>

integration_tests/src/main/python/json_tuple_test.py

@@ -1,4 +1,4 @@
-# Copyright (c) 2023, NVIDIA CORPORATION.
+# Copyright (c) 2023-2024, NVIDIA CORPORATION.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -32,15 +32,17 @@ def test_json_tuple(json_str_pattern):
     assert_gpu_and_cpu_are_equal_collect(
         lambda spark: unary_op_df(spark, gen, length=10).selectExpr(
             'json_tuple(a, "a", "email", "owner", "b", "b$", "b$$")'),
-        conf={'spark.sql.parser.escapedStringLiterals': 'true'})
+        conf={'spark.sql.parser.escapedStringLiterals': 'true',
+            'spark.rapids.sql.expression.JsonTuple': 'true'})
 
 def test_json_tuple_select_non_generator_col():
     gen = StringGen(pattern="{\"Zipcode\":\"abc\",\"ZipCodeType\":\"STANDARD\",\"City\":\"PARC PARQUE\",\"State\":\"PR\"}")
     assert_gpu_and_cpu_are_equal_sql(
         lambda spark : gen_df(spark, [('a', gen)]),
             'table',
             'select a, json_tuple(a, \"Zipcode\", \"ZipCodeType\", \"City\", \"State\") from table',
-        conf={'spark.sql.parser.escapedStringLiterals': 'true'})
+        conf={'spark.sql.parser.escapedStringLiterals': 'true',
+            'spark.rapids.sql.expression.JsonTuple': 'true'})
 
 @allow_non_gpu('GenerateExec', 'JsonTuple')
 @pytest.mark.parametrize('json_str_pattern', json_str_patterns, ids=idfn)
@@ -54,7 +56,8 @@ def test_json_tuple_with_large_number_of_fields_fallback(json_str_pattern):
                            "location", "city", "country", "zip", "code", "region", "state", "street", "block", "loc", \
                            "height", "h", "author", "title", "price", "isbn", "book", "rating", "score", "popular")'),
         "JsonTuple",
-        conf={'spark.sql.parser.escapedStringLiterals': 'true'})
+        conf={'spark.sql.parser.escapedStringLiterals': 'true',
+            'spark.rapids.sql.expression.JsonTuple': 'true'})
 
 @allow_non_gpu('GenerateExec', 'JsonTuple')
 @pytest.mark.parametrize('json_str_pattern', json_str_patterns, ids=idfn)
@@ -66,4 +69,5 @@ def test_json_tuple_with_special_characters_fallback(json_str_pattern):
             lambda spark: unary_op_df(spark, gen, length=10).selectExpr(
                 'json_tuple(a, "a", "a' + special_character + '")'),
             "JsonTuple",
-            conf={'spark.sql.parser.escapedStringLiterals': 'true'})
+            conf={'spark.sql.parser.escapedStringLiterals': 'true',
+                'spark.rapids.sql.expression.JsonTuple': 'true'})

sql-plugin/src/main/scala/com/nvidia/spark/rapids/GpuOverrides.scala

@@ -3733,7 +3733,8 @@ object GpuOverrides extends Logging {
         }
         override def convertToGpu(): GpuExpression = GpuJsonTuple(childExprs.map(_.convertToGpu()))
       }
-    ),
+    ).disabledByDefault("JsonTuple on the GPU does not support all of the normalization " +
+        "that the CPU supports."),
     expr[org.apache.spark.sql.execution.ScalarSubquery](
       "Subquery that will return only one row and one column",
       ExprChecks.projectOnly(

tools/generated_files/supportedExprs.csv

@@ -279,9 +279,9 @@ IsNull,S,`isnull`,None,project,input,S,S,S,S,S,S,S,S,PS,S,S,S,S,NS,PS,PS,PS,NS
 IsNull,S,`isnull`,None,project,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 JsonToStructs,NS,`from_json`,This is disabled by default because it is currently in beta and undergoes continuous enhancements. Please consult the [compatibility documentation](../compatibility.md#json-supporting-types) to determine whether you can enable this configuration for your use case,project,jsonStr,NA,NA,NA,NA,NA,NA,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA
 JsonToStructs,NS,`from_json`,This is disabled by default because it is currently in beta and undergoes continuous enhancements. Please consult the [compatibility documentation](../compatibility.md#json-supporting-types) to determine whether you can enable this configuration for your use case,project,result,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NS,PS,PS,NA
-JsonTuple,S,`json_tuple`,None,project,json,NA,NA,NA,NA,NA,NA,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA
-JsonTuple,S,`json_tuple`,None,project,field,NA,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA
-JsonTuple,S,`json_tuple`,None,project,result,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,S,NA,NA,NA
+JsonTuple,NS,`json_tuple`,This is disabled by default because JsonTuple on the GPU does not support all of the normalization that the CPU supports.,project,json,NA,NA,NA,NA,NA,NA,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA
+JsonTuple,NS,`json_tuple`,This is disabled by default because JsonTuple on the GPU does not support all of the normalization that the CPU supports.,project,field,NA,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA
+JsonTuple,NS,`json_tuple`,This is disabled by default because JsonTuple on the GPU does not support all of the normalization that the CPU supports.,project,result,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,S,NA,NA,NA
 KnownFloatingPointNormalized,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,S,S,S,S,PS,PS,PS,S
 KnownFloatingPointNormalized,S, ,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,S,S,PS,PS,PS,S
 KnownNotNull,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,S,NS,S,S,PS,PS,PS,NS