Skip to content

Latest commit

 

History

History
100 lines (73 loc) · 4.27 KB

CHANGELOG.md

File metadata and controls

100 lines (73 loc) · 4.27 KB

Changelog

SemVer 2.0.0 Keep-A-Changelog 1.0.0

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, and yes, platform and engine support are part of the public API. Please file a bug if you notice a violation of semantic versioning.

Unreleased

Added

Changed

Deprecated

Removed

Fixed

Security

1.0.2 - 2026-02-22

  • TAG: v1.0.2
  • COVERAGE: 98.13% -- 263/268 lines in 10 files
  • BRANCH COVERAGE: 91.18% -- 62/68 branches in 10 files
  • 96.77% documented

Added

  • Benchmarking tools — Performance comparison suite comparing token-resolver against simpler alternatives (String#gsub and Kernel#sprintf):
    • benchmarks/comparison.rb — Comprehensive benchmark script measuring iterations per second across four realistic scenarios (simple replacement, moderate complexity, high complexity, and large documents with sparse tokens)
    • gemfiles/modular/benchmark/ips.gemfile — Development dependency for benchmark-ips gem
    • Rake tasks: rake bench:comparison (run comparison), rake bench:list (list benchmarks), rake bench:run (run all benchmarks), rake bench (alias)
    • BENCHMARK.md — Results and analysis showing token-resolver is 100-3000x slower due to PEG parsing, validation, and AST building; includes guidance on when to use each approach and real-world performance context

1.0.1 - 2026-02-22

  • TAG: v1.0.1
  • COVERAGE: 98.13% -- 263/268 lines in 10 files
  • BRANCH COVERAGE: 91.18% -- 62/68 branches in 10 files
  • 96.77% documented

Added

  • Config#segment_pattern option — a parslet character class constraining which characters are valid inside token segments (default: "[A-Za-z0-9_]"). This prevents false positive token matches against Ruby block parameters ({ |x| expr }), shell variable expansion (${VAR:+val}), and other syntax that structurally resembles tokens but contains spaces or punctuation in the "segments".
  • Resolve#resolve now validates replacement keys against the config's segment_pattern and raises ArgumentError if a key contains characters that the grammar would never parse.

Fixed

  • False positive token matches — the grammar previously used any (match any character) for segment content, which allowed spaces, operators, and punctuation inside token segments. This caused Ruby block syntax like { |fp| File.exist?(fp) } and shell expansion like ${CLASSPATH:+:$CLASSPATH} to be incorrectly parsed as tokens. With multi-separator configs (["|", ":"]), the second | was reconstructed as : during on_missing: :keep roundtripping, silently corrupting source files. The grammar now uses match(segment_pattern) instead of any, limiting segments to word characters by default.

1.0.0 - 2026-02-21

  • TAG: v1.0.0
  • COVERAGE: 97.67% -- 252/258 lines in 10 files
  • BRANCH COVERAGE: 89.39% -- 59/66 branches in 10 files
  • 96.72% documented

Added

  • Initial release

Security