feat: add BIGNUMERIC support to pandas helpers and DB-API

google/cloud/bigquery/_pandas_helpers.py

@@ -81,6 +81,10 @@ def pyarrow_numeric():
     return pyarrow.decimal128(38, 9)
 
 
+def pyarrow_bignumeric():
+    return pyarrow.decimal256(76, 38)
+
+
 def pyarrow_time():
     return pyarrow.time64("us")
 
@@ -104,6 +108,7 @@ def pyarrow_timestamp():
         "INT64": pyarrow.int64,
         "INTEGER": pyarrow.int64,
         "NUMERIC": pyarrow_numeric,
+        "BIGNUMERIC": pyarrow_bignumeric,
         "STRING": pyarrow.string,
         "TIME": pyarrow_time,
         "TIMESTAMP": pyarrow_timestamp,
@@ -132,6 +137,7 @@ def pyarrow_timestamp():
         pyarrow.decimal128(38, scale=9).id: "NUMERIC",
         # The exact decimal's scale and precision are not important, as only
         # the type ID matters, and it's the same for all decimal128 instances.
+        pyarrow.decimal256(76, scale=38).id: "BIGNUMERIC",
     }
 
 else:  # pragma: NO COVER

google/cloud/bigquery/dbapi/_helpers.py

@@ -21,6 +21,11 @@
 
 import six
 
+try:
+    import pyarrow
+except ImportError:  # pragma: NO COVER
+    pyarrow = None
+
 from google.cloud import bigquery
 from google.cloud.bigquery import table
 from google.cloud.bigquery.dbapi import exceptions
@@ -186,7 +191,11 @@ def bigquery_scalar_type(value):
     elif isinstance(value, numbers.Real):
         return "FLOAT64"
     elif isinstance(value, decimal.Decimal):
-        return "NUMERIC"
+        scalar_object = pyarrow.scalar(value)
+        if isinstance(scalar_object, pyarrow.Decimal128Scalar):
+            return "NUMERIC"
+        else:
+            return "BIGNUMERIC"
     elif isinstance(value, six.text_type):
         return "STRING"
     elif isinstance(value, six.binary_type):

google/cloud/bigquery/dbapi/types.py

@@ -78,7 +78,7 @@ def __eq__(self, other):
 STRING = "STRING"
 BINARY = _DBAPITypeObject("BYTES", "RECORD", "STRUCT")
 NUMBER = _DBAPITypeObject(
-    "INTEGER", "INT64", "FLOAT", "FLOAT64", "NUMERIC", "BOOLEAN", "BOOL"
+    "INTEGER", "INT64", "FLOAT", "FLOAT64", "NUMERIC", "BIGNUMERIC", "BOOLEAN", "BOOL"
 )
 DATETIME = _DBAPITypeObject("TIMESTAMP", "DATE", "TIME", "DATETIME")
 ROWID = "ROWID"

google/cloud/bigquery/query.py

@@ -83,7 +83,7 @@ class ScalarQueryParameter(_AbstractQueryParameter):
 
         type_ (str):
             Name of parameter type.  One of 'STRING', 'INT64',
-            'FLOAT64', 'NUMERIC', 'BOOL', 'TIMESTAMP', 'DATETIME', or
+            'FLOAT64', 'NUMERIC', 'BIGNUMERIC', 'BOOL', 'TIMESTAMP', 'DATETIME', or
             'DATE'.
 
         value (Union[str, int, float, decimal.Decimal, bool, datetime.datetime, datetime.date]):
@@ -102,7 +102,7 @@ def positional(cls, type_, value):
         Args:
             type_ (str):
                 Name of parameter type.  One of 'STRING', 'INT64',
-                'FLOAT64', 'NUMERIC', 'BOOL', 'TIMESTAMP', 'DATETIME', or
+                'FLOAT64', 'NUMERIC', 'BIGNUMERIC', 'BOOL', 'TIMESTAMP', 'DATETIME', or
                 'DATE'.
 
             value (Union[str, int, float, decimal.Decimal, bool, datetime.datetime, datetime.date]):
@@ -186,7 +186,7 @@ class ArrayQueryParameter(_AbstractQueryParameter):
 
         array_type (str):
             Name of type of array elements.  One of `'STRING'`, `'INT64'`,
-            `'FLOAT64'`, `'NUMERIC'`, `'BOOL'`, `'TIMESTAMP'`, or `'DATE'`.
+            `'FLOAT64'`, `'NUMERIC'`, `'BIGNUMERIC'`, `'BOOL'`, `'TIMESTAMP'`, or `'DATE'`.
 
         values (List[appropriate scalar type]): The parameter array values.
     """
@@ -203,7 +203,7 @@ def positional(cls, array_type, values):
         Args:
             array_type (str):
                 Name of type of array elements.  One of `'STRING'`, `'INT64'`,
-                `'FLOAT64'`, `'NUMERIC'`, `'BOOL'`, `'TIMESTAMP'`, or `'DATE'`.
+                `'FLOAT64'`, `'NUMERIC'`, `'BIGNUMERIC'`, `'BOOL'`, `'TIMESTAMP'`, or `'DATE'`.
 
             values (List[appropriate scalar type]): The parameter array values.
 

google/cloud/bigquery/schema.py

@@ -32,6 +32,7 @@
     "FLOAT": types.StandardSqlDataType.TypeKind.FLOAT64,
     "FLOAT64": types.StandardSqlDataType.TypeKind.FLOAT64,
     "NUMERIC": types.StandardSqlDataType.TypeKind.NUMERIC,
+    "BIGNUMERIC": types.StandardSqlDataType.TypeKind.BIGNUMERIC,
     "BOOLEAN": types.StandardSqlDataType.TypeKind.BOOL,
     "BOOL": types.StandardSqlDataType.TypeKind.BOOL,
     "GEOGRAPHY": types.StandardSqlDataType.TypeKind.GEOGRAPHY,

tests/unit/test__pandas_helpers.py

@@ -433,6 +433,14 @@ def test_bq_to_arrow_data_type_w_struct_unknown_subfield(module_under_test):
                 "POLYGON ((30 10, 40 40, 20 40, 10 20, 30 10))",
             ],
         ),
+        (
+            "BIGNUMERIC",
+            [
+                decimal.Decimal("-1.123456789012345678901234567890"),
+                None,
+                decimal.Decimal("3.141592653589793238462643383279"),
+            ],
+        ),
     ],
 )
 @pytest.mark.skipif(pandas is None, reason="Requires `pandas`")

tests/unit/test_dbapi__helpers.py

@@ -41,6 +41,10 @@ def test_scalar_to_query_parameter(self):
             (-123456789, "INT64"),
             (1.25, "FLOAT64"),
             (decimal.Decimal("1.25"), "NUMERIC"),
+            (
+                decimal.Decimal("1.1234567890123456789012345678901234567890"),
+                "BIGNUMERIC",
+            ),
             (b"I am some bytes", "BYTES"),
             (u"I am a string", "STRING"),
             (datetime.date(2017, 4, 1), "DATE"),