metadq 0.1.1

Patch release correcting the version reported by the CLI. There are no data-quality behavior changes.

Fixed

• java -jar metadq.jar --version now reports metadq 0.1.1.
• Release smoke tests now verify that the packaged CLI version matches the Git tag on Linux and macOS.

Validation

• Java 17 clean build and test suite
• Linux and macOS artifact smoke tests
• SHA-256 checksum verification
• Zero-data-file demo guard

Limited external validation: The limitations documented for v0.1.0 still apply; feedback from external Iceberg users is welcome.

Full Changelog: v0.1.0...v0.1.1

metadq 0.1.0

metadq is a read-only Java CLI that runs conservative data-quality checks from Apache Iceberg metadata—without Spark or content data-file scans.

Limited external validation: This release passed CI, Linux and macOS artifact checks, checksum verification, and the local MinIO/REST end-to-end test, but it has not yet been validated by external Iceberg users.

Quick start

Java 17 or newer is required.

curl -LO https://github.com/keemgdeok/metadq/releases/download/v0.1.0/metadq.jar
java -jar metadq.jar doctor --demo

Included

• doctor explains which metadata evidence is available.
• check evaluates table.row_count, column.null_ratio, table.last_commit_age, and schema.column.
• Text and JSON output with documented process exit codes.
• REST catalog and S3/S3-compatible storage support.
• A reproducible MinIO + Iceberg REST Catalog environment.

The documented Apple M4 benchmark collected metadata for 100,000 live files in 75 ms for row count, 81 ms with one requested column, and 127 ms with all ten columns. These are in-memory metadata timings, not a comparison with a data-row scan.

Validation

• The exact release JAR passed its checksum and demo guard on Ubuntu and macOS with Java 17.
• The MinIO/REST end-to-end test produced four rows and passed all four rule families.
• GitHub Actions CI and the guarded release workflow completed successfully.

Known limits

• Results are only as complete as the metrics written into Iceberg metadata.
• Incomplete metrics or applicable delete files can produce UNKNOWN.
• Iceberg format v1 and v2 are supported; v3 is rejected.
• A managed REST catalog with S3 has not yet been validated.

Feedback wanted

Please open an issue if UNKNOWN is unclear, catalog setup fails, or you identify a real table or workflow where you would use metadq.

Changes since v0.1.0-rc.1

• Improve README readability and project overview
• Validate release artifacts on Linux and macOS
• Release metadq 0.1.0

Full changelog: v0.1.0-rc.1...v0.1.0

metadq 0.1.0-rc.1

metadq is a read-only Java CLI that runs conservative data-quality checks from Apache Iceberg metadata—without Spark or content data-file scans.

Limited external validation: This release candidate passed CI, Linux and macOS artifact checks, checksum verification, and the local MinIO/REST end-to-end test, but it has not yet been validated by external Iceberg users.

Quick start

Java 17 or newer is required.

curl -LO https://github.com/keemgdeok/metadq/releases/download/v0.1.0-rc.1/metadq.jar
java -jar metadq.jar doctor --demo

Included

• doctor explains which metadata evidence is available.
• check evaluates table.row_count, column.null_ratio, table.last_commit_age, and schema.column.
• Text and JSON output with documented process exit codes.
• REST catalog and S3/S3-compatible storage support.
• A reproducible MinIO + Iceberg REST Catalog environment.

The documented Apple M4 benchmark collected metadata for 100,000 live files in 75 ms for row count, 81 ms with one requested column, and 127 ms with all ten columns. These are in-memory metadata timings, not a comparison with a data-row scan.

Known limits

• Results are only as complete as the metrics written into Iceberg metadata.
• Incomplete metrics or applicable delete files can produce UNKNOWN.
• Iceberg format v1 and v2 are supported; v3 is rejected.
• A managed REST catalog with S3 has not yet been validated.

Feedback wanted

Please open an issue if UNKNOWN is unclear, catalog setup fails, or you identify a real table or workflow where you would use metadq.

Changes

• Harden Iceberg evidence collection and add reproducible validation
• Prepare 0.1.0 release candidate

Full changelog: https://github.com/keemgdeok/metadq/commits/v0.1.0-rc.1