GH-3511: Add JMH encoding benchmarks and fix parquet-benchmarks shaded jar by iemejia · Pull Request #3512 · apache/parquet-java

GH-3511: Add JMH encoding benchmarks and fix parquet-benchmarks shaded jar

iemejia

committed

GH-3511: Isolate setup cost and benchmark full dictionary paths

iemejia

committed

GH-3511: Tighten dictionary benchmarks and reduce setup coupling

iemejia

committed

Add JMH CompressionBenchmark for isolated codec throughput measurement

iemejia

committed

Fine-tune JMH benchmarks for targeted per-branch measurement

iemejia

committed

GH-3522: Add batch read APIs to ValuesReader hierarchy (from par13)

iemejia

committed

Override readIntegers() in RLE ValuesReader to delegate to batch decoder (from par13)

iemejia

committed

Add ValuesReader-level and batch RLE decode benchmarks

iemejia

committed

Add PLAIN decoding benchmark for all numeric types (INT32/INT64/FLOAT/DOUBLE)

iemejia

committed

Add batch decode benchmarks to ByteStreamSplitDecodingBenchmark

iemejia

committed

Add BOOLEAN encoding characterization benchmark (V1 PLAIN vs V2 RLE)

iemejia

committed

Add batch decode benchmarks to BooleanEncodingBenchmark (V1 PLAIN + V2 RLE)

iemejia

committed

Add dictionary encoding benchmark for LONG, FLOAT, DOUBLE, and FLBA types

iemejia

committed

Expand FLBA benchmark: add DELTA_BYTE_ARRAY, BSS, DICTIONARY encode+decode

iemejia

committed

Add LZ4_RAW to file-level benchmark @Param codec lists

iemejia

committed

Add batch encode benchmarks for BOOLEAN writeBooleans() API

iemejia

committed

Add batch encode benchmarks for PLAIN numeric encoding (INT32/INT64/FLOAT/DOUBLE)

iemejia

committed

Add FLBA PLAIN batch encode/decode benchmarks (encodePlainBatch, decodePlainBatch)

iemejia

committed

Add BROTLI to CompressionBenchmark codec parameter list

iemejia

committed

Add BROTLI direct bypass in DirectCodecFactory using jbrotli JNI

iemejia

committed

Use bulk ByteBuffer view reads in BSS batch decode for all numeric types

iemejia

committed

Optimize BOOLEAN decoding: direct bit extraction and batch readBooleans()

iemejia

committed

Use lookup table for boolean batch decode (PLAIN + RLE PACKED)

iemejia

committed

Optimize BOOLEAN encoding: batch writeBooleans() with direct byte packing

iemejia

committed

Optimize PLAIN encoding: batch writeIntegers/writeLongs/writeFloats/writeDoubles with bulk ByteBuffer view transfers

iemejia

committed

Add batch read/write APIs for fixed-length byte arrays

iemejia

committed

Optimize FLBA BSS encoding: batched scatter replacing per-byte stream writes

iemejia

committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

GH-3511: Add JMH encoding benchmarks and fix parquet-benchmarks shaded jar#3512

GH-3511: Add JMH encoding benchmarks and fix parquet-benchmarks shaded jar#3512
iemejia wants to merge 27 commits into
apache:masterfrom
iemejia:perf-benchmarks

Commits on May 13, 2026

GH-3511: Add JMH encoding benchmarks and fix parquet-benchmarks shaded jar

GH-3511: Isolate setup cost and benchmark full dictionary paths

GH-3511: Tighten dictionary benchmarks and reduce setup coupling

Add JMH CompressionBenchmark for isolated codec throughput measurement

Fine-tune JMH benchmarks for targeted per-branch measurement

GH-3522: Add batch read APIs to ValuesReader hierarchy (from par13)

Override readIntegers() in RLE ValuesReader to delegate to batch decoder (from par13)

Add ValuesReader-level and batch RLE decode benchmarks

Add PLAIN decoding benchmark for all numeric types (INT32/INT64/FLOAT/DOUBLE)

Add batch decode benchmarks to ByteStreamSplitDecodingBenchmark

Add BOOLEAN encoding characterization benchmark (V1 PLAIN vs V2 RLE)

Add batch decode benchmarks to BooleanEncodingBenchmark (V1 PLAIN + V2 RLE)

Add dictionary encoding benchmark for LONG, FLOAT, DOUBLE, and FLBA types

Expand FLBA benchmark: add DELTA_BYTE_ARRAY, BSS, DICTIONARY encode+decode

Add LZ4_RAW to file-level benchmark @Param codec lists

Add batch encode benchmarks for BOOLEAN writeBooleans() API

Add batch encode benchmarks for PLAIN numeric encoding (INT32/INT64/FLOAT/DOUBLE)

Add FLBA PLAIN batch encode/decode benchmarks (encodePlainBatch, decodePlainBatch)

Add BROTLI to CompressionBenchmark codec parameter list

Add BROTLI direct bypass in DirectCodecFactory using jbrotli JNI

Use bulk ByteBuffer view reads in BSS batch decode for all numeric types

Optimize BOOLEAN decoding: direct bit extraction and batch readBooleans()

Use lookup table for boolean batch decode (PLAIN + RLE PACKED)

Optimize BOOLEAN encoding: batch writeBooleans() with direct byte packing

Optimize PLAIN encoding: batch writeIntegers/writeLongs/writeFloats/writeDoubles with bulk ByteBuffer view transfers

Add batch read/write APIs for fixed-length byte arrays

Optimize FLBA BSS encoding: batched scatter replacing per-byte stream writes