Hardwood - modern lightweight Java reader and writer for Parquet file format

Gunnar Morling released Hardwood 1.1.0.Beta2 on 2026-10-12, adding row group control in the writer, faster S3 reads and footer caching to the lightweight Java library and command-line tool for reading and writing Apache Parquet files. Hardwood, which reached 1.0 on 2026-06-25, has zero transitive dependencies beyond optional compression libraries (Snappy, ZSTD, LZ4, Brotli) and aims to be the fastest Parquet reader and writer for the JVM without pulling in Hadoop, Avro or the parquet-java dependency tree. The 1.1.0.Beta1 release on 2026-08-31 added Parquet write support and Bloom filter pruning, and the project ships a drop-in shim module for migration from parquet-java plus a CLI skill that lets agents inspect and analyse Parquet files.

Hardwood ¶ hardwood A modern lightweight Java reader and writer for the Apache Parquet https://parquet.apache.org/ file format. Available as a Java library and a command-line tool. What's new - 2026-10-12 — Hardwood 1.1.0.Beta2 https://www.morling.dev/blog/row-group-control-faster-s3-reads-footer-caching-hardwood-1-1-0-beta2/ : row group control in the writer, faster S3 reads, footer caching - 2026-08-31 — Hardwood 1.1.0.Beta1 https://www.morling.dev/blog/parquet-file-write-support-bloom-filters-improved-performance-hardwood-1-1-0-beta1/ : Parquet write support, Bloom filter pruning, faster reads - 2026-06-25 — Hardwood 1.0 https://www.morling.dev/blog/hardwood-1-0-fast-lightweight-apache-parquet-reader-for-the-jvm/ : the release announcement and the story behind the project - 2026-06-01 — Confluent Developer Podcast, Ep. 31 https://www.youtube.com/watch?v=9Ov0Cn ArHE : Gunnar Morling on building a new Parquet parser with AI Why Hardwood ¶ why-hardwood Hardwood gives applications fast and efficient support for reading and writing Parquet, without pulling in Hadoop, Avro, or the wider parquet-java https://github.com/apache/parquet-java dependency tree. It is built to be: - Light-weight : Zero transitive dependencies beyond optional compression libraries Snappy, ZSTD, LZ4, Brotli - Fast : Hardwood aims to be the fastest Parquet reader and writer for the JVM 1.0 read benchmarks https://www.morling.dev/blog/hardwood-1-0-fast-lightweight-apache-parquet-reader-for-the-jvm/ performance - Complete : Read and write support for flat and nested schemas, every logical type, every primitive type in current use, and the encodings and codecs in current use, with new format additions tracked as they land - Scalable : Hardwood is multi-threaded at the core, pages are decoded in parallel, with cross-file prefetching for multi-file reads - Embeddable : The Hardwood library can be used in GraalVM native binaries; a WASM build of dive runs in the browser as a preview https://hardwood.dev/experiments/dive-web/ - Agent-friendly : Hardwood's CLI comes with a skill which lets your agents inspect and analyse Parquet files - Compatible : A drop-in shim module how-to/compat/ facilitates migration from parquet-java , with documented divergences where Hardwood applies stricter semantics e.g. SQL three-valued notEq The CLI reference/cli/ inspects and analyses Parquet files and includes an interactive TUI for exploring a file's schema, row groups, pages, and data. Quick Example ¶ quick-example Reading a file with the row-based API how-to/row-reader/ : python import dev.hardwood.InputFile; import dev.hardwood.reader.ParquetFileReader; import dev.hardwood.reader.RowReader; try ParquetFileReader fileReader = ParquetFileReader.open InputFile.of path ; RowReader rowReader = fileReader.rowReader { while rowReader.hasNext { rowReader.next ; long id = rowReader.getLong "id" ; String name = rowReader.getString "name" ; LocalDate birthDate = rowReader.getDate "birth date" ; Instant createdAt = rowReader.getTimestamp "created at" ; } } Writing a file how-to/write-row-by-row/ row by row: python import dev.hardwood.OutputFile; import dev.hardwood.writer.ParquetFileWriter; import dev.hardwood.writer.RowWriter; try ParquetFileWriter writer = ParquetFileWriter.create OutputFile.of path , schema { RowWriter rows = writer.rowWriter ; for Person person : people { rows.writeRow row - row .setLong "id", person.id .setString "name", person.name .setDate "birth date", person.birthDate ; } } Set up Hardwood in your project getting-started/ , then read your first file end-to-end tutorial/first-read/ . Alternatively, install the Hardwood CLI getting-started/ command-line-tool for working with Parquet files on the command line. Status and Limitations ¶ status-and-limitations The read path is stable and suitable for production use; ColumnReader , with its Validity and LayerKind types, is annotated @Experimental , so its shape may still change. The write path ships both APIs described under Choosing a Writer how-to/ choosing-a-writer ; the row-oriented RowWriter and its builders, the ColumnBatch setters for nested and nullable columns, and the in-memory destination InMemoryOutputFile are annotated @Experimental , so their shape may still change. Writing targets local files through OutputFile.of Path and the heap through OutputFile.inMemory . Reads have no file-size ceiling, but a column chunk read without an OffsetIndex is capped at 2 GB, and some backends carry their own limits; see Limits reference/reader/ limits . Roadmap ¶ roadmap Forward-looking items, none committed to a specific release. - Finalize ColumnReader API — stabilize the API for columnar access and move it out of "Experimental" state. 522 https://github.com/hardwood-hq/hardwood/issues/522 - Writer extensions — object-store output and parallel column encoding, on top of the write path described in The Write Model concepts/write-model/ . 9 https://github.com/hardwood-hq/hardwood/issues/9 - Parquet Modular Encryption — read files encrypted under the Parquet Modular Encryption spec https://github.com/apache/parquet-format/blob/master/Encryption.md : encrypted footer, per-column keys, AES-GCM and AES-GCM-CTR. 128 https://github.com/hardwood-hq/hardwood/issues/128 - Apache Arrow interop — ColumnReader output as Arrow FieldVector / VectorSchemaRoot for zero-copy handoff to DuckDB, DataFusion, Pandas-via-JNI, and other Arrow-native consumers. 153 https://github.com/hardwood-hq/hardwood/issues/153 Getting help ¶ getting-help Ask questions on GitHub Discussions https://github.com/hardwood-hq/hardwood/discussions and file bugs on the issue tracker https://github.com/hardwood-hq/hardwood/issues ; see Contributing contributing/ . Articles, talks & podcasts ¶ articles-talks-podcasts - Hardwood: A New Parser for Apache Parquet https://www.morling.dev/blog/hardwood-new-parser-for-apache-parquet/ original project announcement - Building a New Parquet Engine with AI https://www.youtube.com/watch?v=9Ov0Cn ArHE Confluent Developer podcast - Hardwood Promises High-Speed JVM Apache Parquet Processing with Zero Mandatory Dependencies https://www.infoq.com/news/2026/07/hardwood-java-parquet/ InfoQ article - Hardwood: Building a Parquet Parser From Scratch With a Little Help From AI https://speakerdeck.com/gunnarmorling/hardwood-building-a-parquet-parser-from-scratch-with-a-little-help-from-ai conference talk - GitHub Open Source Friday with Gunnar Morling https://www.youtube.com/watch?v=teqFSSQEtCw GitHub podcast - Chasing Efficient Java Development: From 1BRC to Developing Hardwood AI Natively https://www.infoq.com/podcasts/chasing-efficient-java-development/ InfoQ podcast