Skip to content

Commit db97c39

Browse files
anthony-tuiningaanthony-tuininga
committed
Added support for the NCHAR, NVARCHAR and NCLOB data types in data
frames (#505).
1 parent 3cc4270 commit db97c39

File tree

5 files changed

+78
-2
lines changed

5 files changed

+78
-2
lines changed

‎doc/src/release_notes.rst‎

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -69,6 +69,8 @@ Common Changes
6969
(`issue 499 <https://github.com/oracle/python-oracledb/issues/499>`__).
7070
- Fixed bug on Windows when fetching dates prior to 1970 and after 2038
7171
(`issue 483 <https://github.com/oracle/python-oracledb/issues/483>`__).
72+
- Added support for the NCHAR, NVARCHAR and NCLOB data types
73+
(`issue 505 <https://github.com/oracle/python-oracledb/issues/505>`__).
7274

7375
#) Added parameter ``pool_name`` to connection and pool creation methods to
7476
support Oracle Database 23ai multi-pool :ref:`drcp`.

‎src/oracledb/impl/base/metadata.pyx‎

Lines changed: 4 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -71,7 +71,8 @@ cdef class OracleMetadata:
7171
self._arrow_type = NANOARROW_TYPE_INT64
7272
else:
7373
self._arrow_type = NANOARROW_TYPE_DOUBLE
74-
elif db_type_num in (DB_TYPE_NUM_CHAR, DB_TYPE_NUM_VARCHAR):
74+
elif db_type_num in (DB_TYPE_NUM_CHAR, DB_TYPE_NUM_VARCHAR,
75+
DB_TYPE_NUM_NCHAR, DB_TYPE_NUM_NVARCHAR):
7576
self._arrow_type = NANOARROW_TYPE_STRING
7677
elif db_type_num == DB_TYPE_NUM_BINARY_FLOAT:
7778
self._arrow_type = NANOARROW_TYPE_FLOAT
@@ -86,7 +87,8 @@ cdef class OracleMetadata:
8687
self._arrow_type = NANOARROW_TYPE_TIMESTAMP
8788
elif db_type_num == DB_TYPE_NUM_LONG_RAW:
8889
self._arrow_type = NANOARROW_TYPE_LARGE_BINARY
89-
elif db_type_num == DB_TYPE_NUM_LONG_VARCHAR:
90+
elif db_type_num in (DB_TYPE_NUM_LONG_VARCHAR,
91+
DB_TYPE_NUM_LONG_NVARCHAR):
9092
self._arrow_type = NANOARROW_TYPE_LARGE_STRING
9193
elif db_type_num == DB_TYPE_NUM_RAW:
9294
self._arrow_type = NANOARROW_TYPE_BINARY

‎src/oracledb/impl/thick/var.pyx‎

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -369,8 +369,11 @@ cdef class ThickVarImpl(BaseVarImpl):
369369
ora_data.buffer.as_bool = data.value.asBoolean
370370
elif ora_type_num in (
371371
DPI_ORACLE_TYPE_CHAR,
372+
DPI_ORACLE_TYPE_LONG_NVARCHAR,
372373
DPI_ORACLE_TYPE_LONG_VARCHAR,
373374
DPI_ORACLE_TYPE_LONG_RAW,
375+
DPI_ORACLE_TYPE_NCHAR,
376+
DPI_ORACLE_TYPE_NVARCHAR,
374377
DPI_ORACLE_TYPE_RAW,
375378
DPI_ORACLE_TYPE_VARCHAR,
376379
):

‎tests/test_8000_dataframe.py‎

Lines changed: 8 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -683,11 +683,15 @@ def test_8028(self):
683683
cast(1.23 as binary_double) as col_double,
684684
cast(7.14 as binary_float) as col_float,
685685
cast('abcd' as varchar2(10)) as col_string,
686+
cast('efgh' as nvarchar2(6)) as col_nstring,
687+
cast('ijkl' as char(4)) as col_char,
688+
cast('mnop' as nchar(4)) as col_nchar,
686689
cast(systimestamp as timestamp(0)) as col_ts_sec,
687690
cast(systimestamp as timestamp(3)) as col_ts_ms,
688691
cast(systimestamp as timestamp(6)) as col_ts_us,
689692
cast(systimestamp as timestamp(9)) as col_ts_ns,
690693
to_clob('abc') as col_large_string,
694+
to_nclob('def') as col_large_nstring,
691695
utl_raw.cast_to_raw('abc2') as col_binary,
692696
to_blob(utl_raw.cast_to_raw('abc3')) as col_large_binary
693697
from dual
@@ -706,11 +710,15 @@ def test_8028(self):
706710
"COL_DOUBLE": (dtype_kind.FLOAT, 64, "g", "="),
707711
"COL_FLOAT": (dtype_kind.FLOAT, 64, "g", "="),
708712
"COL_STRING": (dtype_kind.STRING, 8, "u", "="),
713+
"COL_NSTRING": (dtype_kind.STRING, 8, "u", "="),
714+
"COL_CHAR": (dtype_kind.STRING, 8, "u", "="),
715+
"COL_NCHAR": (dtype_kind.STRING, 8, "u", "="),
709716
"COL_TS_SEC": (dtype_kind.DATETIME, 64, "tss:", "="),
710717
"COL_TS_MS": (dtype_kind.DATETIME, 64, "tsm:", "="),
711718
"COL_TS_US": (dtype_kind.DATETIME, 64, "tsu:", "="),
712719
"COL_TS_NS": (dtype_kind.DATETIME, 64, "tsn:", "="),
713720
"COL_LARGE_STRING": (dtype_kind.STRING, 8, "U", "="),
721+
"COL_LARGE_NSTRING": (dtype_kind.STRING, 8, "U", "="),
714722
"COL_BINARY": (dtype_kind.STRING, 8, "z", "="),
715723
"COL_LARGE_BINARY": (dtype_kind.STRING, 8, "Z", "="),
716724
"COL_DECIMAL128": (dtype_kind.DECIMAL, 128, "d:10.2", "="),

‎tests/test_8100_dataframe_async.py‎

Lines changed: 61 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -628,6 +628,67 @@ async def test_8123(self):
628628
fetched_data = self.__get_data_from_df(fetched_df)
629629
self.assertEqual(fetched_data, data)
630630

631+
async def test_8124(self):
632+
"8124 - verify dtype for all Arrow types"
633+
query = """
634+
select
635+
cast(1 as number(10)) as col_int64,
636+
cast(1.23 as binary_double) as col_double,
637+
cast(7.14 as binary_float) as col_float,
638+
cast('abcd' as varchar2(10)) as col_string,
639+
cast('efgh' as nvarchar2(6)) as col_nstring,
640+
cast('ijkl' as char(4)) as col_char,
641+
cast('mnop' as nchar(4)) as col_nchar,
642+
cast(systimestamp as timestamp(0)) as col_ts_sec,
643+
cast(systimestamp as timestamp(3)) as col_ts_ms,
644+
cast(systimestamp as timestamp(6)) as col_ts_us,
645+
cast(systimestamp as timestamp(9)) as col_ts_ns,
646+
to_clob('abc') as col_large_string,
647+
to_nclob('def') as col_large_nstring,
648+
utl_raw.cast_to_raw('abc2') as col_binary,
649+
to_blob(utl_raw.cast_to_raw('abc3')) as col_large_binary
650+
from dual
651+
"""
652+
decimal_query = (
653+
"select cast(123.45 as decimal(10, 2)) as col_decimal128"
654+
)
655+
656+
# determine dtype kind enumeration
657+
ora_df = await self.conn.fetch_df_all("select user from dual")
658+
col = ora_df.get_column(0)
659+
dtype_kind = type(col.dtype[0])
660+
661+
expected_dtypes = {
662+
"COL_INT64": (dtype_kind.INT, 64, "l", "="),
663+
"COL_DOUBLE": (dtype_kind.FLOAT, 64, "g", "="),
664+
"COL_FLOAT": (dtype_kind.FLOAT, 64, "g", "="),
665+
"COL_STRING": (dtype_kind.STRING, 8, "u", "="),
666+
"COL_NSTRING": (dtype_kind.STRING, 8, "u", "="),
667+
"COL_CHAR": (dtype_kind.STRING, 8, "u", "="),
668+
"COL_NCHAR": (dtype_kind.STRING, 8, "u", "="),
669+
"COL_TS_SEC": (dtype_kind.DATETIME, 64, "tss:", "="),
670+
"COL_TS_MS": (dtype_kind.DATETIME, 64, "tsm:", "="),
671+
"COL_TS_US": (dtype_kind.DATETIME, 64, "tsu:", "="),
672+
"COL_TS_NS": (dtype_kind.DATETIME, 64, "tsn:", "="),
673+
"COL_LARGE_STRING": (dtype_kind.STRING, 8, "U", "="),
674+
"COL_LARGE_NSTRING": (dtype_kind.STRING, 8, "U", "="),
675+
"COL_BINARY": (dtype_kind.STRING, 8, "z", "="),
676+
"COL_LARGE_BINARY": (dtype_kind.STRING, 8, "Z", "="),
677+
"COL_DECIMAL128": (dtype_kind.DECIMAL, 128, "d:10.2", "="),
678+
}
679+
680+
# check query without fetch_decimals enabled
681+
ora_df = await self.conn.fetch_df_all(query)
682+
for i, name in enumerate(ora_df.column_names()):
683+
col = ora_df.get_column(i)
684+
self.assertEqual(col.dtype, expected_dtypes[name])
685+
686+
# check query with fetch_decimals enabled
687+
with test_env.DefaultsContextManager("fetch_decimals", True):
688+
ora_df = await self.conn.fetch_df_all(decimal_query)
689+
col = ora_df.get_column(0)
690+
self.assertEqual(col.dtype, expected_dtypes["COL_DECIMAL128"])
691+
631692

632693
if __name__ == "__main__":
633694
test_env.run_test_cases()

0 commit comments

Comments
 (0)