v0.4.1

Xulu CLI 0.4.1

Date: 2026-09-14

Added

• xulu profile
  • Profile a CSV, JSON, JSONL, or Parquet file: types, nulls/completeness, uniqueness, min/max or lengths, and formats
  • Text and JSON reports via --report / -r (text default), with --pretty and --color
  • Optional --sample-rows, --json-root, --delimiter, --data-format, --decimals, and --sort

v0.4.0

Xulu CLI 0.4.0

Date: 2026-09-14

Added

• xulu schema lint
  • Check if a schema file is valid (YAML or JSON)
  • Reports unknown / not-yet-available fields and parse errors
  • Text and JSON reports via --report / -r (text default), with --pretty and --color

v0.3.2

Xulu CLI 0.3.2

Date: 2026-09-13

Changed

• xulu inspect
  • Clearer detailed text structure for JSON / JSONL; with --color, keys and counts stay soft gray and type names are bright white

v0.3.1

Xulu CLI 0.3.1

Date: 2026-09-09

Added

• xulu diff
  • --ignore-field <name> — ignore matching CSV, TSV, SSV, PSV, or Parquet columns and JSON / JSONL fields at any depth
  • --ignore-path <path> — ignore a selected JSON / JSONL field or subtree using paths with an optional $, array indexes, * / [*], or **
  • Both options can be repeated and work with positional, Myers, and key-based comparisons; fields used by --key cannot also be ignored

v0.3.0

Xulu CLI 0.3.0

Date: 2026-09-06

Added

Xulu now supports a complete local schema workflow: generate a schema from your data, review it, and validate files locally or in CI.

• xulu validate
  • Validate CSV, TSV, SSV, PSV, JSON, JSONL, NDJSON, and Parquet against a YAML or JSON schema.
  • Check types, required values, bounds, lengths, enums, formats, uniqueness, null ratios, table shape, and nested JSON fields.
  • Read clear terminal or JSON reports and use validation results in CI.
• xulu schema generate
  • Generate a starting YAML or JSON schema from CSV, TSV, SSV, PSV, JSON, JSONL, NDJSON, or Parquet.
  • Review and edit the generated rules, then use the schema with xulu validate.

Changed

• xulu diff
  • Dropped column indexes in text reports for better readability: ~ row[2].column[2] (name) → ~ row[2].name (JSON reports still include indexes)
• xulu inspect
  • --depth 0 now means unlimited nesting for detailed JSON / JSONL (same idea as --max-keys 0)

v0.2.13

Xulu CLI 0.2.13

Date: 2026-09-05

Added

• Windows x86_64 support

v0.2.12

Xulu CLI 0.2.12

Date: 2026-09-05

Added

• macOS Apple Silicon support

v0.2.11

Xulu CLI 0.2.11

Date: 2026-09-04

Added

• xulu diff
  • -a key / --algorithm key — match rows by column instead of position (order does not matter)
  • --key SPEC — join column(s), e.g. --key id, --key id,date, or --key id=user_id
  • --ignore-added — with -a key, check that left is contained in right (ignore extras only on the right)

Changed

• xulu diff
  • Clearer error when -a key is used without --key (includes examples)

Removed

• xulu diff
  • Long aliases --absolute-tolerance and --relative-tolerance (use --atol / --rtol)

v0.2.10

Xulu CLI 0.2.10

Date: 2026-09-03

Added

• CLI
  • Unknown flags and values suggest the closest match
• xulu diff
  • Support for absolute and relative numeric tolerances (--atol / --absolute-tolerance, --rtol / --relative-tolerance)
  • Support for type coercion (--coerce-types), with independent string-as-number and empty-as-zero policies
  • Equivalence rules are shown in text and JSON reports

v0.2.9

Xulu CLI 0.2.9

Date: 2026-09-03

Added

• xulu update
  • Linux arm64 self-update (downloads xulu-linux-aarch64, same as install)

Fixed

• xulu diff
  • Mixed-format pairs (for example Parquet vs CSV) now say both sides must use the same format, instead of incorrectly claiming one format is unsupported