[SPARK-52249][PS] Enable divide-by-zero for numeric truediv with ANSI enabled

### What changes were proposed in this pull request?
Enable divide-by-zero for truediv with ANSI enabled

### Why are the changes needed?
Part of https://issues.apache.org/jira/browse/SPARK-52169

### Does this PR introduce _any_ user-facing change?
Yes,  divide-by-zero for truediv is enabled with ANSI enabled

```py
>>> spark.conf.get("spark.sql.ansi.enabled")
'true'
>>> pdf = pd.DataFrame({"a": [1.0, -1.0, 0.0, np.nan], "b": [0.0, 0.0, 0.0, 0.0]})
>>> psdf = ps.from_pandas(pdf)
```

FROM
```py
>>> psdf["a"] / psdf["b"]
...
pyspark.errors.exceptions.captured.ArithmeticException: [DIVIDE_BY_ZERO] Division by zero. Use `try_divide` to tolerate divisor being 0 and return NULL instead. If necessary set "spark.sql.ansi.enabled" to "false" to bypass this error. SQLSTATE: 22012
== DataFrame ==
"__div__" was called from
<stdin>:1
```

TO
```py
>>> psdf["a"] / psdf["b"]
0    inf
1   -inf
2    NaN
3    NaN
dtype: float64
```

### How was this patch tested?
Unit tests

### Was this patch authored or co-authored using generative AI tooling?
No

Closes #50972 from xinrong-meng/divide_0.

Authored-by: Xinrong Meng <xinrong@apache.org>
Signed-off-by: Takuya Ueshin <ueshin@databricks.com>

Author: xinrong-meng

python/pyspark/pandas/data_type_ops/num_ops.py

@@ -43,6 +43,7 @@
     _is_boolean_type,
 )
 from pyspark.pandas.typedef.typehints import extension_dtypes, pandas_on_spark_type
+from pyspark.pandas.utils import is_ansi_mode_enabled
 from pyspark.sql import functions as F, Column as PySparkColumn
 from pyspark.sql.types import (
     BooleanType,
@@ -247,14 +248,23 @@ def truediv(self, left: IndexOpsLike, right: Any) -> SeriesOrIndex:
         _sanitize_list_like(right)
         if not is_valid_operand_for_numeric_arithmetic(right):
             raise TypeError("True division can not be applied to given types.")
+        spark_session = left._internal.spark_frame.sparkSession
+        right = transform_boolean_operand_to_numeric(right, spark_type=left.spark.data_type)
 
         def truediv(left: PySparkColumn, right: Any) -> PySparkColumn:
-            return F.when(
-                F.lit(right != 0) | F.lit(right).isNull(),
-                left.__div__(right),
-            ).otherwise(F.lit(np.inf).__div__(left))
+            if is_ansi_mode_enabled(spark_session):
+                return F.when(
+                    F.lit(right == 0),
+                    F.when(left < 0, F.lit(float("-inf")))
+                    .when(left > 0, F.lit(float("inf")))
+                    .otherwise(F.lit(np.nan)),
+                ).otherwise(left / right)
+            else:
+                return F.when(
+                    F.lit(right != 0) | F.lit(right).isNull(),
+                    left.__div__(right),
+                ).otherwise(F.lit(np.inf).__div__(left))
 
-        right = transform_boolean_operand_to_numeric(right, spark_type=left.spark.data_type)
         return numpy_column_op(truediv)(left, right)
 
     def floordiv(self, left: IndexOpsLike, right: Any) -> SeriesOrIndex:
@@ -332,18 +342,27 @@ def truediv(self, left: IndexOpsLike, right: Any) -> SeriesOrIndex:
         _sanitize_list_like(right)
         if not is_valid_operand_for_numeric_arithmetic(right):
             raise TypeError("True division can not be applied to given types.")
+        spark_session = left._internal.spark_frame.sparkSession
+        right = transform_boolean_operand_to_numeric(right, spark_type=left.spark.data_type)
 
         def truediv(left: PySparkColumn, right: Any) -> PySparkColumn:
-            return F.when(
-                F.lit(right != 0) | F.lit(right).isNull(),
-                left.__div__(right),
-            ).otherwise(
-                F.when(F.lit(left == np.inf) | F.lit(left == -np.inf), left).otherwise(
-                    F.lit(np.inf).__div__(left)
+            if is_ansi_mode_enabled(spark_session):
+                return F.when(
+                    F.lit(right == 0),
+                    F.when(left < 0, F.lit(float("-inf")))
+                    .when(left > 0, F.lit(float("inf")))
+                    .otherwise(F.lit(np.nan)),
+                ).otherwise(left / right)
+            else:
+                return F.when(
+                    F.lit(right != 0) | F.lit(right).isNull(),
+                    left.__div__(right),
+                ).otherwise(
+                    F.when(F.lit(left == np.inf) | F.lit(left == -np.inf), left).otherwise(
+                        F.lit(np.inf).__div__(left)
+                    )
                 )
-            )
 
-        right = transform_boolean_operand_to_numeric(right, spark_type=left.spark.data_type)
         return numpy_column_op(truediv)(left, right)
 
     def floordiv(self, left: IndexOpsLike, right: Any) -> SeriesOrIndex:

python/pyspark/pandas/tests/computation/test_binary_ops.py

@@ -111,7 +111,6 @@ def test_binary_operator_sub(self):
         psdf = ps.DataFrame({"a": ["x"], "b": ["y"]})
         self.assertRaisesRegex(TypeError, ks_err_msg, lambda: psdf["a"] - psdf["b"])
 
-    @unittest.skipIf(is_ansi_mode_test, ansi_mode_not_supported_message)
     def test_divide_by_zero_behavior(self):
         # float / float
         # np.float32

python/pyspark/pandas/utils.py

@@ -1070,6 +1070,13 @@ def xor(df1: PySparkDataFrame, df2: PySparkDataFrame) -> PySparkDataFrame:
     )
 
 
+def is_ansi_mode_enabled(spark: SparkSession) -> bool:
+    return (
+        ps.get_option("compute.ansi_mode_support", spark_session=spark)
+        and spark.conf.get("spark.sql.ansi.enabled") == "true"
+    )
+
+
 def _test() -> None:
     import os
     import doctest