[SPARK-39392][SQL][3.3] Refine ANSI error messages for try_* function…

… hints

### What changes were proposed in this pull request?

Refine ANSI error messages and remove 'To return NULL instead'.
This PR is a backport of #36780 from `master`

### Why are the changes needed?

Improve error messaging for ANSI mode since the user may not even aware that query was returning NULLs.

### Does this PR introduce _any_ user-facing change?

No

### How was this patch tested?

Unit tests

Closes #36792 from vli-databricks/SPARK-39392-3.3.

Authored-by: Vitalii Li <vitalii.li@databricks.com>
Signed-off-by: Max Gekk <max.gekk@gmail.com>

core/src/main/resources/error/error-classes.json

@@ -26,11 +26,11 @@
     "message" : [ "Cannot use a mixture of aggregate function and group aggregate pandas UDF" ]
   },
   "CAST_INVALID_INPUT" : {
-    "message" : [ "The value <value> of the type <sourceType> cannot be cast to <targetType> because it is malformed. Correct the value as per the syntax, or change its target type. To return NULL instead, use `try_cast`. If necessary set <config> to \"false\" to bypass this error." ],
+    "message" : [ "The value <value> of the type <sourceType> cannot be cast to <targetType> because it is malformed. Correct the value as per the syntax, or change its target type. Use `try_cast` to tolerate malformed input and return NULL instead. If necessary set <config> to \"false\" to bypass this error." ],
     "sqlState" : "42000"
   },
   "CAST_OVERFLOW" : {
-    "message" : [ "The value <value> of the type <sourceType> cannot be cast to <targetType> due to an overflow. To return NULL instead, use `try_cast`. If necessary set <config> to \"false\" to bypass this error." ],
+    "message" : [ "The value <value> of the type <sourceType> cannot be cast to <targetType> due to an overflow. Use `try_cast` to tolerate overflow and return NULL instead. If necessary set <config> to \"false\" to bypass this error." ],
     "sqlState" : "22005"
   },
   "CONCURRENT_QUERY" : {
@@ -41,7 +41,7 @@
     "sqlState" : "22008"
   },
   "DIVIDE_BY_ZERO" : {
-    "message" : [ "Division by zero. To return NULL instead, use `try_divide`. If necessary set <config> to \"false\" (except for ANSI interval type) to bypass this error." ],
+    "message" : [ "Division by zero. Use `try_divide` to tolerate divisor being 0 and return NULL instead. If necessary set <config> to \"false\" (except for ANSI interval type) to bypass this error." ],
     "sqlState" : "22012"
   },
   "DUPLICATE_KEY" : {
@@ -93,7 +93,7 @@
     "message" : [ "The index <indexValue> is out of bounds. The array has <arraySize> elements. If necessary set <config> to \"false\" to bypass this error." ]
   },
   "INVALID_ARRAY_INDEX_IN_ELEMENT_AT" : {
-    "message" : [ "The index <indexValue> is out of bounds. The array has <arraySize> elements. To return NULL instead, use `try_element_at`. If necessary set <config> to \"false\" to bypass this error." ]
+    "message" : [ "The index <indexValue> is out of bounds. The array has <arraySize> elements. Use `try_element_at` to tolerate accessing element at invalid index and return NULL instead. If necessary set <config> to \"false\" to bypass this error." ]
   },
   "INVALID_FIELD_NAME" : {
     "message" : [ "Field name <fieldName> is invalid: <path> is not a struct." ],
@@ -115,7 +115,7 @@
     "sqlState" : "42000"
   },
   "MAP_KEY_DOES_NOT_EXIST" : {
-    "message" : [ "Key <keyValue> does not exist. To return NULL instead, use `try_element_at`. If necessary set <config> to \"false\" to bypass this error." ]
+    "message" : [ "Key <keyValue> does not exist. Use `try_element_at` to tolerate non-existent key and return NULL instead. If necessary set <config> to \"false\" to bypass this error." ]
   },
   "MISSING_COLUMN" : {
     "message" : [ "Column '<columnName>' does not exist. Did you mean one of the following? [<proposal>]" ],

core/src/test/scala/org/apache/spark/SparkThrowableSuite.scala

@@ -125,7 +125,8 @@ class SparkThrowableSuite extends SparkFunSuite {
 
     // Does not fail with too many args (expects 0 args)
     assert(getMessage("DIVIDE_BY_ZERO", Array("foo", "bar", "baz")) ==
-      "Division by zero. To return NULL instead, use `try_divide`. If necessary set foo " +
+      "Division by zero. Use `try_divide` to tolerate divisor being 0 and return NULL instead. " +
+        "If necessary set foo " +
       "to \"false\" (except for ANSI interval type) to bypass this error.")
   }
 

sql/catalyst/src/main/scala/org/apache/spark/sql/errors/QueryExecutionErrors.scala

@@ -493,7 +493,9 @@ private[sql] object QueryExecutionErrors extends QueryErrorsBase {
       message: String,
       hint: String = "",
       errorContext: String = ""): ArithmeticException = {
-    val alternative = if (hint.nonEmpty) s" To return NULL instead, use '$hint'." else ""
+    val alternative = if (hint.nonEmpty) {
+      s" Use '$hint' to tolerate overflow and return NULL instead."
+    } else ""
     new SparkArithmeticException(
       errorClass = "ARITHMETIC_OVERFLOW",
       messageParameters = Array(message, alternative, SQLConf.ANSI_ENABLED.key),
@@ -1093,7 +1095,8 @@ private[sql] object QueryExecutionErrors extends QueryErrorsBase {
       value: Any, from: DataType, to: DataType, errorContext: String): Throwable = {
     val valueString = toSQLValue(value, from)
     new DateTimeException(s"Invalid input syntax for type ${toSQLType(to)}: $valueString. " +
-      s"To return NULL instead, use 'try_cast'. If necessary set ${SQLConf.ANSI_ENABLED.key} " +
+      s"Use `try_cast` to tolerate malformed input and return NULL instead. " +
+      s"If necessary set ${SQLConf.ANSI_ENABLED.key} " +
       s"to false to bypass this error." + errorContext)
   }
 

sql/core/src/test/resources/sql-tests/results/ansi/array.sql.out

@@ -168,7 +168,7 @@ select element_at(array(1, 2, 3), 5)
 struct<>
 -- !query output
 org.apache.spark.SparkArrayIndexOutOfBoundsException
-The index 5 is out of bounds. The array has 3 elements. To return NULL instead, use `try_element_at`. If necessary set "spark.sql.ansi.enabled" to "false" to bypass this error.
+The index 5 is out of bounds. The array has 3 elements. Use `try_element_at` to tolerate accessing element at invalid index and return NULL instead. If necessary set "spark.sql.ansi.enabled" to "false" to bypass this error.
 
 
 -- !query
@@ -177,7 +177,7 @@ select element_at(array(1, 2, 3), -5)
 struct<>
 -- !query output
 org.apache.spark.SparkArrayIndexOutOfBoundsException
-The index -5 is out of bounds. The array has 3 elements. To return NULL instead, use `try_element_at`. If necessary set "spark.sql.ansi.enabled" to "false" to bypass this error.
+The index -5 is out of bounds. The array has 3 elements. Use `try_element_at` to tolerate accessing element at invalid index and return NULL instead. If necessary set "spark.sql.ansi.enabled" to "false" to bypass this error.
 
 
 -- !query
@@ -337,7 +337,7 @@ select element_at(array(1, 2, 3), 5)
 struct<>
 -- !query output
 org.apache.spark.SparkArrayIndexOutOfBoundsException
-The index 5 is out of bounds. The array has 3 elements. To return NULL instead, use `try_element_at`. If necessary set "spark.sql.ansi.enabled" to "false" to bypass this error.
+The index 5 is out of bounds. The array has 3 elements. Use `try_element_at` to tolerate accessing element at invalid index and return NULL instead. If necessary set "spark.sql.ansi.enabled" to "false" to bypass this error.
 
 
 -- !query
@@ -346,7 +346,7 @@ select element_at(array(1, 2, 3), -5)
 struct<>
 -- !query output
 org.apache.spark.SparkArrayIndexOutOfBoundsException
-The index -5 is out of bounds. The array has 3 elements. To return NULL instead, use `try_element_at`. If necessary set "spark.sql.ansi.enabled" to "false" to bypass this error.
+The index -5 is out of bounds. The array has 3 elements. Use `try_element_at` to tolerate accessing element at invalid index and return NULL instead. If necessary set "spark.sql.ansi.enabled" to "false" to bypass this error.
 
 
 -- !query