rok commented on code in PR #36701:
URL: https://github.com/apache/arrow/pull/36701#discussion_r1975761191
##########
python/pyarrow/tests/test_array.py:
##########
@@ -3286,24 +3295,34 @@ def test_numpy_binary_overflow_to_chunked():
unicode_values += [unicode_unique_strings[i % 10]
for i in range(1 << 11)]
- for case, ex_type in [(values, pa.binary()),
- (unicode_values, pa.utf8())]:
+ binary_type = pa.large_binary() if large_types else pa.binary()
+ string_type = pa.large_utf8() if large_types else pa.utf8()
+ for case, ex_type in [(values, binary_type),
+ (unicode_values, string_type)]:
arr = np.array(case)
- arrow_arr = pa.array(arr)
+ arrow_arr = pa.array(arr, ex_type)
arr = None
- assert isinstance(arrow_arr, pa.ChunkedArray)
assert arrow_arr.type == ex_type
+ if large_types:
+ # Large types shouldn't be chunked
+ assert isinstance(arrow_arr, pa.Array)
+
+ for i in range(len(arrow_arr)):
+ val = arrow_arr[i]
+ assert val.as_py() == case[i]
+ else:
+ assert isinstance(arrow_arr, pa.ChunkedArray)
- # Split up into 16MB chunks. 128 * 16 = 2048, so 129
- assert arrow_arr.num_chunks == 129
+ # Split up into 16MB chunks. 128 * 16 = 2048, so 129
+ assert arrow_arr.num_chunks == 129
- value_index = 0
- for i in range(arrow_arr.num_chunks):
- chunk = arrow_arr.chunk(i)
- for val in chunk:
- assert val.as_py() == case[value_index]
- value_index += 1
+ value_index = 0
+ for i in range(arrow_arr.num_chunks):
+ chunk = arrow_arr.chunk(i)
+ for val in chunk:
+ assert val.as_py() == case[value_index]
+ value_index += 1
Review Comment:
Oh sorry, I didn't read the diff right.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
