Repository navigation
[python] Normalize nested naive timestamps when writing Avro to object stores and HDFS - #10356
Merged
JingsongLi merged 1 commit intoOct 26, 2026
Conversation
Contributor
|
Please rebase master. |
Akash3121
suggested changes
Oct 25, 2026
Akash3121
left a comment
Akash3121
left a comment
Contributor
There was a problem hiding this comment.
@jackylee-ch Rebase to master, code looks good.
jackylee-ch
force-pushed
the
python-avro-nested-timestamp-objectstore
branch
from
October 26, 2026 00:41
328b761 to
df5b0f4
Compare
…t stores and HDFS LocalFileIO.write_avro recursively attaches UTC to naive datetimes nested in ROW/ARRAY/MAP values, but PyArrowFileIO (OSS/S3/GCS) and HdfsNativeFileIO normalized only top-level columns. On a non-UTC host, fastavro then interprets a naive datetime nested in a struct or list using the host timezone, so the persisted instant is shifted by the host offset -- silent data corruption on the object-store and HDFS Avro write paths that the local path already handled. Move the recursive normalizer to FileIO (common) and use it in all three write_avro implementations. Add a PyArrowFileIO round-trip test alongside the existing LocalFileIO one, both asserting a nested TIMESTAMP survives a write under TZ=Asia/Shanghai.
jackylee-ch
force-pushed
the
python-avro-nested-timestamp-objectstore
branch
from
October 26, 2026 08:02
df5b0f4 to
2797968
Compare
Contributor
|
+1 |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Purpose
LocalFileIO.write_avrorecursively attaches UTC to naive datetimes nested in ROW/ARRAY/MAP values, butPyArrowFileIO(OSS/S3/GCS) andHdfsNativeFileIOnormalized only top-level column values. On a non-UTC host, fastavro interprets a naivedatetimeusing the host timezone, so a naiveTIMESTAMPnested in a struct or list is persisted shifted by the host offset — silent data corruption on the object-store and HDFS Avro write paths that the local path already handled.This moves the recursive normalizer onto
FileIOand uses it in all threewrite_avroimplementations.Tests
data_types_test.py: the existingtest_avro_nested_timestamp_roundtrips_on_non_utc_host(LocalFileIO) now exercises the shared helper; a newtest_pyarrow_file_io_avro_normalizes_nested_naive_timestampsasserts a nestedTIMESTAMPsurvives aPyArrowFileIO.write_avroround-trip underTZ=Asia/Shanghai. Disabling the recursion fails both.API and Format
No change.
Documentation
No change.