Repository navigation
Conversation
|
Thanks for opening a pull request! This pull request has been automatically converted to a draft because its title doesn't match Arrow's required format. If this is not a minor PR. Could you open an issue for this pull request on GitHub? https://github.com/apache/arrow/issues/new/choose Opening GitHub issues ahead of time contributes to the Openness of the Apache Arrow project. Then could you also rename the pull request title in the following format? or After updating the title, you can mark the pull request as ready for review. See also: |
github-actions
Bot
added
awaiting review
Add the FSST sources, license, notice, and release-audit metadata required to build Parquet support from a pinned in-tree implementation.
Encode BYTE_ARRAY pages with FSST and serialize exported decoder tables and lengths in a bounded little-endian layout. Add reader, writer, metadata, Python, malformed-input, and round-trip coverage for the experimental encoding.
Train bounded token dictionaries, greedily encode strings, and decode rows at arbitrary offsets. Preserve binary input, configurable dictionary budgets, bit-packed codes, and explicit bounds for malformed inputs and output buffers.
Split each string into a group prefix and suffix before FSST or OnPair sees the data. Keep reconstruction row-addressable and cover empty groups, binary input, and boundaries between prefix groups.
Measure projected size, encoding, full-column decoding, and row access for FSST, OnPair, and prefix variants over shared generated corpora. Validate decoded bytes before timing and keep parameter selection outside reported codec work.
Compare native Parquet pages and projected FSST and OnPair payloads with optional generic compression, including dictionary-then-OnPair. Separate parameter selection from timed encoding and verify every decoded column.
Exercise dictionary budgets from 9 through 16 and isolate packed-code width from dictionary contents in a dedicated benchmark. Fail on any round-trip mismatch and cover each remainder left by grouped decoding.
Build a decode-only dictionary with fixed token slots so each lookup needs one random payload access. Leave the compact stored dictionary unchanged and retain portable, short-stream, and caller-prebuilt paths.
Unroll packed decoding by a multiple of each width's byte-alignment period so offsets and shifts remain constant within a group. Preserve scalar tails and cover every possible group remainder.
Train FSST with sparse pair counts and 16-byte symbols, then encode through the shared OnPair parser and decoder. This isolates table construction while keeping tokenization, packing, and row reconstruction identical.
Seed literal tokens only for bytes present in the column and dispatch every resulting code width. Preserve complete token coverage, bounded dictionary capacity, and round trips for narrow alphabets.
Train candidate budgets and choose the smallest projected decode time within a bounded compression regression. Validate policy inputs, deterministic ties, fallbacks, and selection across corpus shapes.
Validate configurations, offsets, tokens, dictionaries, packed streams, selector policies, and output capacities before use. Propagate explicit sizes through every caller and retain malformed-input, sanitizer, and remainder coverage.
Compile the maintained OnPair support library, five standalone tests, and two benchmarks through Arrow's build. Keep raw fitting programs on the research branch and leave the evaluation library uninstalled.
prtkgaur
force-pushed
the
pgaurOnPairEvaluation
branch
2 times, most recently
from
October 4, 2026 19:35
880c425 to
53149d6
Compare
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
3 participants
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.
Thanks for opening a pull request!
If this is your first pull request you can find detailed information on how to contribute here:
Please remove this line and the above text before creating your pull request.
Rationale for this change
What changes are included in this PR?
Are these changes tested?
Are there any user-facing changes?
This PR includes breaking changes to public APIs. (If there are any breaking changes to public APIs, please explain which changes are breaking. If not, you can remove this.)
This PR contains a "Critical Fix". (If the changes fix either (a) a security vulnerability, (b) a bug that caused incorrect or invalid data to be produced, or (c) a bug that causes a crash (even when the API contract is upheld), please provide explanation. If not, you can remove this.)