[copy_from] Support the Parquet format #31173
Open
Conversation
This file contains 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
* plumb the CSV params through to oneshot_source::CsvFormat to handle different delimiters and what not
* add support for compressed CSVs using async_compression
* various Bazel changes for the C dependencies from the compression algorithms
* add a new OneshotSource implementation * support the FILES and PATTERN options for COPY FROM
* Add a new Parquet OneshotFormat * Fix ranged requests for HTTP and AWS sources
5 tasks
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Outdated suggestions cannot be applied.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.
Stacked on top of #31144
This PR adds a new implementation of a
OneshotFormatthat supports reading in Parquet files. The decoding is built on top of theArrowReaderimplemented in #30958.The strategy we use for reading and decoding Parquet files is the "split work" stage of a oneshot source will read the footer metadata from a Parquet file to determine the Row Group boundaries. The Row Groups are then distributed among timely works for fetching and eventual decoding.
Note: Through experimentation I found that Row Groups seem to typically be 10s of MB large, which makes them a pretty good unit of parallelization.
Motivation
Fixes https://github.com/MaterializeInc/database-issues/issues/8853
Tips for reviewer
Review on the final commit, the one titled "start, support Parquet for COPY FROM"
Checklist
$T ⇔ Proto$Tmapping (possibly in a backwards-incompatible way), then it is tagged with aT-protolabel.