Zoltan Ivanfi created PARQUET-1628:
--------------------------------------
Summary: Accept local timestamps annotated with the legacy
timestamp types
Key: PARQUET-1628
URL: https://issues.apache.org/jira/browse/PARQUET-1628
Project: Parquet
Issue Type: Task
Components: parquet-mr
Reporter: Zoltan Ivanfi
Assignee: Nandor Kollar
The rules for TIMESTAMP forward-compatibility were created based on the
assumption that TIMESTAMP_MILLIS and TIMESTAMP_MICROS have only been used in
the instant aka. UTC-normalized semantics so far.
>From this false premise it followed that TIMESTAMPs with local semantics were
>a new type and did not need to be annotated with the old types to maintain
>compatibility. In fact, annotating them with the old types were considered to
>be harmful, since it would have mislead older readers into thinking that they
>can read TIMESTAMPs with local semantics, when in reality they would have
>misinterpreted them as TIMESTAMPs with instant semantics. This would have lead
>to a difference of several hours, corresponding to the time zone offset.
In reality, however, this misinterpretation of timestamps has already been
going on for a while, since Arrow annotates local timestamps with
TIMESTAMP_MILLIS or TIMESTMAP_MICROS.
To maintain forward compatibilty of local timestamps, Arrow annotates them with
the legacy timestamp logical types. However, the Java library considers these
logical types to be incompatible and discards the new type in favour of the
legacy ones (since doing the other way around would change the behaviour).
Parquet-mr should be updated so that it accepts this combination of new and old
logical types.
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)