Changelog

View Source

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

Unreleased

0.3.0 - 2025-11-03

100% TOON Specification v1.3.3 Compliance Achieved

Tested against official fixtures from toon-format/spec@b9c71f7 (v1.3.3, 2025-10-31)

Changed

  • BREAKING: Switched to official TOON specification fixtures for testing
  • Removed all custom test files in favor of official test fixtures (306 tests)
  • Added git submodule for toon-format/spec
  • Removed unused lib/toon/decode/strings.ex (functionality merged into StructuralParser)

Added

  • Complete decoder implementation with 100% specification compliance (160/160 tests)
  • Strict mode validation for decoder (indentation, blank lines, array lengths)
  • Support for root primitive values (single strings, numbers, etc.)
  • Length validation for inline, tabular, and list arrays
  • Custom indent_size option for decoder
  • Quoted string handling with delimiter-aware splitting
  • Leading zero number handling (treats "05" as string per spec)
  • Invalid escape sequence detection
  • Unterminated string detection
  • Support for # length marker prefix (e.g., [#3]:)
  • Empty list item support (- → {})
  • Nested arrays in list items
  • Tabular arrays with quoted keys containing delimiters
  • Comprehensive escape sequence validation

Fixed

  • String escaping now handles all five valid escapes (\\, \", \n, \r, \t)
  • Quoted values in delimited arrays now properly unescaped
  • Root primitive detection now handles quoted strings with colons
  • Nested object encoding in list items now has correct indentation
  • Decoder correctly handles commas in object values when not in an array
  • Array length mismatches now properly raise errors in strict mode
  • Tabular row count validation
  • Custom delimiter support in nested arrays and tabular formats
  • Float precision increased to 17 digits for full round-trip fidelity

Testing

  • Test assertions modified to compare semantic equivalence (decode both outputs and compare)
  • This approach validates correctness independent of key ordering differences in Elixir 1.19
  • All 306 official TOON specification tests now pass (100% compliance)

Critical Bug Fixes

  • Fixed nested field parsing in list items (indent level tracking)
  • Fixed tabular array data row indentation in nested contexts
  • Fixed list array item indentation when nested in other list items
  • Fixed take/skip logic for array data vs. sibling fields

0.2.0 - 2025-10-28

Fixed

  • Tabular array data rows now properly indented at depth + 1
  • List-style array items now properly indented
  • Top-level arrays no longer include spurious "items[N]:" header
  • README examples updated to match actual output

Added

  • Comprehensive doctests for all modules (110 doctests total)
  • 12 new doctest files covering encode, decode, shared, and error modules
  • Integration tests for array indentation at all nesting levels

0.1.0 - 2025-10-28

Added

  • Initial implementation of TOON encoder and decoder for Elixir
  • Full TOON format support (primitives, objects, arrays)
  • Three array formats: inline, tabular, and list
  • Toon.Encoder protocol for custom struct encoding
  • Comprehensive type specifications with Dialyzer support
  • Telemetry instrumentation for encoding and decoding operations
  • Property-based testing with StreamData
  • Complete documentation with examples
  • Benchmarks comparing TOON vs JSON token efficiency