tustvold opened a new pull request, #4376:
URL: https://github.com/apache/arrow-rs/pull/4376
_Draft as needs a bit more polish_
# Which issue does this PR close?
<!--
We generally require a GitHub issue to be filed for all bug fixes and
enhancements and this helps us generate change logs for our releases. You can
link an issue to this PR using the GitHub syntax. For example `Closes #123`
indicates that this PR will close issue #123.
-->
Closes #4365
# Rationale for this change
<!--
Why are you proposing this change? If this is already explained clearly in
the issue then this section is not needed.
Explaining clearly why changes are proposed helps reviewers understand your
changes and offer better suggestions for fixes.
-->
Previously when reading repeated fields, RecordReader would read in batches
of at least 1024 levels. From this it would then split off the levels
comprising the desired number of rows. Aside from being wasteful, and complex,
this method causes issues for selection pushdown as it would need to skip into
data that has already potentially been decoded.
# What changes are included in this PR?
<!--
There is no need to duplicate the description in the issue here but it is
sometimes worth providing a summary of the individual changes in this PR.
-->
This changes ColumnReader to delimit records, whilst this does potentially
impose overheads on users using this API directly, I struggle to contrive a
use-case where reading partial records would be desirable / correct. I
therefore hypothesize such users are either not using repeated data (likely) or
are already performing some sort of record buffering similar to what is now
performed as a first-party API.
# Are there any user-facing changes?
Technically no, however, marking this as breaking as it does alter the
behaviour of ColumnReader, especially with regards to ill-formed level data as
one might encounter in a test.
<!--
If there are user-facing changes then we may require documentation to be
updated before approving the PR.
-->
<!---
If there are any breaking changes to public APIs, please add the `breaking
change` label.
-->
--
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]
