This is an automated email from the ASF dual-hosted git repository.
paleolimbot pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/arrow-nanoarrow.git
The following commit(s) were added to refs/heads/main by this push:
new 761ead52 fix(Python) Use timezone-aware timestamps in some unit tests
on Windows (#512)
761ead52 is described below
commit 761ead522ed94998c4c9001249c52ea0c141e20a
Author: Abhishek Singh <[email protected]>
AuthorDate: Sun Jun 9 18:05:50 2024 -0700
fix(Python) Use timezone-aware timestamps in some unit tests on Windows
(#512)
Using `datetime.timestamp()` for timestamps close to EPOCH fails on
Windows crashing Python.
Please check this run
https://github.com/apache/arrow-nanoarrow/actions/runs/9428246621/job/25973355558
Following snippet
```python
def test_c_array_timestamp_microseconds():
d1 = int(round(datetime(1970, 1, 1).timestamp() * 1e6))
```
raises error
```bash
OSError: [Errno 22] Invalid argument
```
The crux of the problem is explained in this [CPython
issue](https://github.com/python/cpython/issues/81708). Solution is to
use timezone aware datetime objects
---------
Co-authored-by: Dewey Dunnington <[email protected]>
---
python/tests/test_c_array.py | 36 ++++++++++++++++++------------------
python/tests/test_c_buffer.py | 8 ++++----
2 files changed, 22 insertions(+), 22 deletions(-)
diff --git a/python/tests/test_c_array.py b/python/tests/test_c_array.py
index 2db9bdb8..6df96472 100644
--- a/python/tests/test_c_array.py
+++ b/python/tests/test_c_array.py
@@ -16,7 +16,7 @@
# under the License.
import array
-from datetime import date, datetime
+from datetime import date, datetime, timezone
import pytest
from nanoarrow._lib import CArrayBuilder, NanoarrowException
@@ -520,9 +520,9 @@ def test_c_array_from_buffers_validation():
def test_c_array_timestamp_seconds():
- d1 = int(round(datetime(1970, 1, 1).timestamp()))
- d2 = int(round(datetime(1985, 12, 31).timestamp()))
- d3 = int(round(datetime(2005, 3, 4).timestamp()))
+ d1 = int(round(datetime(1970, 1, 1, tzinfo=timezone.utc).timestamp()))
+ d2 = int(round(datetime(1985, 12, 31, tzinfo=timezone.utc).timestamp()))
+ d3 = int(round(datetime(2005, 3, 4, tzinfo=timezone.utc).timestamp()))
c_array = na.c_array([d1, d2, d3], na.timestamp("s"))
assert c_array.length == 3
assert c_array.null_count == 0
@@ -532,9 +532,9 @@ def test_c_array_timestamp_seconds():
def test_c_array_timestamp_seconds_from_pybuffer():
- d1 = int(round(datetime(1970, 1, 1).timestamp()))
- d2 = int(round(datetime(1985, 12, 31).timestamp()))
- d3 = int(round(datetime(2005, 3, 4).timestamp()))
+ d1 = int(round(datetime(1970, 1, 1, tzinfo=timezone.utc).timestamp()))
+ d2 = int(round(datetime(1985, 12, 31, tzinfo=timezone.utc).timestamp()))
+ d3 = int(round(datetime(2005, 3, 4, tzinfo=timezone.utc).timestamp()))
c_array = na.c_array(array.array("q", [d1, d2, d3]), na.timestamp("s"))
assert c_array.length == 3
assert c_array.null_count == 0
@@ -544,9 +544,9 @@ def test_c_array_timestamp_seconds_from_pybuffer():
def test_c_array_timestamp_milliseconds():
- d1 = int(round(datetime(1970, 1, 1).timestamp() * 1e3))
- d2 = int(round(datetime(1985, 12, 31).timestamp() * 1e3))
- d3 = int(round(datetime(2005, 3, 4).timestamp() * 1e3))
+ d1 = int(round(datetime(1970, 1, 1, tzinfo=timezone.utc).timestamp() *
1e3))
+ d2 = int(round(datetime(1985, 12, 31, tzinfo=timezone.utc).timestamp() *
1e3))
+ d3 = int(round(datetime(2005, 3, 4, tzinfo=timezone.utc).timestamp() *
1e3))
c_array = na.c_array([d1, d2, d3], na.timestamp("ms"))
assert c_array.length == 3
assert c_array.null_count == 0
@@ -556,9 +556,9 @@ def test_c_array_timestamp_milliseconds():
def test_c_array_timestamp_milliseconds_from_pybuffer():
- d1 = int(round(datetime(1970, 1, 1).timestamp() * 1e3))
- d2 = int(round(datetime(1985, 12, 31).timestamp() * 1e3))
- d3 = int(round(datetime(2005, 3, 4).timestamp() * 1e3))
+ d1 = int(round(datetime(1970, 1, 1, tzinfo=timezone.utc).timestamp() *
1e3))
+ d2 = int(round(datetime(1985, 12, 31, tzinfo=timezone.utc).timestamp() *
1e3))
+ d3 = int(round(datetime(2005, 3, 4, tzinfo=timezone.utc).timestamp() *
1e3))
c_array = na.c_array(array.array("q", [d1, d2, d3]), na.timestamp("ms"))
assert c_array.length == 3
assert c_array.null_count == 0
@@ -568,9 +568,9 @@ def test_c_array_timestamp_milliseconds_from_pybuffer():
def test_c_array_timestamp_microseconds():
- d1 = int(round(datetime(1970, 1, 1).timestamp() * 1e6))
- d2 = int(round(datetime(1985, 12, 31).timestamp() * 1e6))
- d3 = int(round(datetime(2005, 3, 4).timestamp() * 1e6))
+ d1 = int(round(datetime(1970, 1, 1, tzinfo=timezone.utc).timestamp() *
1e6))
+ d2 = int(round(datetime(1985, 12, 31, tzinfo=timezone.utc).timestamp() *
1e6))
+ d3 = int(round(datetime(2005, 3, 4, tzinfo=timezone.utc).timestamp() *
1e6))
c_array = na.c_array([d1, d2, d3], na.timestamp("us"))
assert c_array.length == 3
assert c_array.null_count == 0
@@ -580,8 +580,8 @@ def test_c_array_timestamp_microseconds():
def test_c_array_timestamp_nanoseconds():
- d1 = int(round(datetime(1970, 1, 1).timestamp() * 1e9))
- d2 = int(round(datetime(1985, 12, 31).timestamp() * 1e9))
+ d1 = int(round(datetime(1970, 1, 1, tzinfo=timezone.utc).timestamp() *
1e9))
+ d2 = int(round(datetime(1985, 12, 31, tzinfo=timezone.utc).timestamp() *
1e9))
d3 = int(round(datetime(2005, 3, 4).timestamp() * 1e9))
c_array = na.c_array([d1, d2, d3], na.timestamp("ns"))
assert c_array.length == 3
diff --git a/python/tests/test_c_buffer.py b/python/tests/test_c_buffer.py
index 94111d5f..fbd5538d 100644
--- a/python/tests/test_c_buffer.py
+++ b/python/tests/test_c_buffer.py
@@ -17,7 +17,7 @@
import struct
import sys
-from datetime import date, datetime
+from datetime import date, datetime, timezone
import pytest
from nanoarrow._lib import CBuffer, CBufferBuilder
@@ -366,9 +366,9 @@ def test_c_buffer_bitmap_from_iterable():
def test_c_buffer_from_timestamp_iterable():
- d1 = int(round(datetime(1970, 1, 1).timestamp() * 1e3))
- d2 = int(round(datetime(1985, 12, 31).timestamp() * 1e3))
- d3 = int(round(datetime(2005, 3, 4).timestamp() * 1e3))
+ d1 = int(round(datetime(1970, 1, 1, tzinfo=timezone.utc).timestamp() *
1e3))
+ d2 = int(round(datetime(1985, 12, 31, tzinfo=timezone.utc).timestamp() *
1e3))
+ d3 = int(round(datetime(2005, 3, 4, tzinfo=timezone.utc).timestamp() *
1e3))
with pytest.raises(ValueError):
na.c_buffer([d1, d2, d3], na.timestamp("ms"))
