[MINOR][VL] Remove dead Arrow-CSV / Arrow-Dataset JVM code paths#12130
Open
yaooqinn wants to merge 2 commits into
Open
[MINOR][VL] Remove dead Arrow-CSV / Arrow-Dataset JVM code paths#12130yaooqinn wants to merge 2 commits into
yaooqinn wants to merge 2 commits into
Conversation
The ArrowCSV file format and ArrowBatchScanExec chain are unreachable: no injection in VeloxRuleApi, no META-INF/services entry, and all ArrowCsvScanSuite cases are @ignore'd. They were introduced as a squash-merge byproduct in apache#11776 and never wired up. Verified by compiling: * spark-3.5 + scala-2.12 + arrow 15.0.0-gluten (install) * spark-4.0 + scala-2.13 + arrow 18.1.0 (compile) Generated-by: claude-opus-4.7
makeArrowDiscovery / readArrowSchema / readArrowFileColumnNames / readSchema(FragmentScanOptions) overloads / loadMissingColumns / loadPartitionColumns / loadBatch in ArrowUtil have zero callers across the repo after the previous removal of the ArrowCSV chain. Drop them together with the now-unused imports (arrow.dataset.*, FileStatus, URI/URLDecoder, ArrowRecordBatch, Optional, Logging, etc.). Verified by compiling: * spark-3.5 + scala-2.12 (test-compile, patched arrow 15.0.0-gluten) * spark-4.0 + scala-2.13 (compile, pure arrow 18.1.0) Generated-by: claude-opus-4.7
|
Run Gluten Clickhouse CI on x86 |
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What changes were proposed in this pull request?
Removes dead JVM-side code paths related to Arrow CSV scanning and Arrow Dataset readers in
gluten-arrow. Two commits:[MINOR][VL] Remove dead Arrow-CSV / Arrow-Dataset JVM code path(-1441 lines)ArrowCSVFileFormat,ArrowCSVOptionConverter,ArrowCSVPartitionReaderFactory,ArrowCSVScan,ArrowCSVScanBuilder,ArrowCSVTable,ArrowBatchScanExec,ArrowConvertorRule,ArrowScanReplaceRule,ArrowFileSourceScanExec,BaseArrowScanExec,ArrowCsvScanSuite.ArrowBatchScanExecShimsegment from 5 shim files (spark33/34/35/40/41).[MINOR][VL] Remove dead Arrow dataset reader paths from ArrowUtil(-167 lines)ArrowUtil.scala:makeArrowDiscovery,readArrowSchema,readArrowFileColumnNames,readSchema(×2 overloads),loadMissingColumns,loadPartitionColumns,loadBatch. These all instantiatedFileSystemDatasetFactory/CsvFragmentScanOptionsand were only used by the now-deleted classes above.Total: 18 files changed, -1608 lines.
Why are the changes needed?
This code is dead:
META-INF/servicesentry routes Spark toArrowCSVFileFormatorArrowCSVTable.VeloxRuleApidoes not injectArrowConvertorRuleorArrowScanReplaceRuleinto the optimizer pipeline.ArrowCsvScanSuiteis fully@Ignored.greping the whole repo for the 6ArrowUtilreader methods returns 0 call sites outside the deleted files.These classes appear to be unreachable code introduced by a previous squash-merge and never wired into the actual execution path. They also keep
gluten-arrowglued to the patchedarrow-datasetJVM API (CsvFragmentScanOptions.from(Map), 5-argFileSystemDatasetFactory) shipped viadev/build-arrow.sh— removing them unblocks future work to drop the patched-arrow build entirely.Does this PR introduce any user-facing change?
No.
How was this patch tested?
Compiled clean on both Spark profiles:
All 6 affected modules pass scalastyle with 0 errors.
Generated-by: claude-opus-4.7