Skip to content

CHANGELOG.md

Latest commit

 

History

History
103 lines (71 loc) · 4.04 KB

CHANGELOG.md

File metadata and controls

103 lines (71 loc) · 4.04 KB
Raw

Changelog

1.0.0

Major Changes

  • 92468f7: BREAKING CHANGES: Migrate to Bun and add standalone binary compilation

    This release introduces significant infrastructure changes:

    • ESM-only package: Removed CommonJS builds. The package now exports only ESM format.
    • Bun development requirement: Development now requires Bun 1.2+ instead of pnpm.
    • Node.js runtime compatibility maintained: Runtime consumption still supports Node.js 22+.

    New Features:

    • Standalone binary compilation using Bun's --compile flag
      • Cross-platform support: linux-x64, linux-arm64, darwin-x64, darwin-arm64, windows-x64
      • Baseline and modern CPU variants (Haswell/AVX2+)
      • Automated GitHub workflow for binary builds on releases
    • Runtime detection for Bun.CryptoHasher with Node.js fallback for hash computation

    Build System:

    • Replaced tsup with Bun.build() for faster builds (~0.6s vs 1.5s)
    • Simplified package exports to ESM default only
    • Updated all 30+ benchmark scripts to use Bun
    • Removed pnpm workspace configuration

0.5.1

Patch Changes

  • 5cf4806: - Guard Readability extraction against fragment HTML by reparsing missing <html> wrappers, fixing documentElement access errors.
    • Inline sample article fixtures in tests to avoid Git LFS pointers and keep CI checks working out of the box.

0.5.0

Minor Changes

  • 4541105: Refined the htmlparser2 renderer to stream nodes directly, which removes the domhandler dependency and unlocks faster parse times on large fixtures. Expose granular convert_parse, convert_render, and convert_postprocess telemetry events (propagated through H2MParser and htmlToMarkdown) so downstream tooling can pinpoint hotspots while staying backwards compatible.

0.4.0

Minor Changes

  • 9a25c34: Optimize package size and update Node.js requirements

    • Reduce npm package size by 82% (570KB → 103KB) by excluding source maps and enabling minification
    • Remove unused turndown dependency, moving it to devDependencies for benchmarks only
    • Update Node.js requirement to >=22 and test matrix to Node 22 and 24
    • Fix tsup config TypeScript error in esbuildOptions
    • Update all GitHub Actions workflows to use Node.js 24
    • Configure package.json to exclude source maps from published package
    • Enable minification in production builds while keeping it disabled for watch mode

    These changes improve installation speed and reduce disk usage for end users while maintaining full functionality.

0.3.0

Minor Changes

  • 5e93d99: Fast HTML to Markdown converter with Mozilla Readability extraction, streaming renderer, and LLM-ready output. 3.75x faster than alternatives

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

0.1.0 - 2025-01-27

Added

  • Initial release of h2m-parser
  • Fast HTML to Markdown conversion using htmlparser2
  • Mozilla Readability integration for article extraction
  • Streaming Markdown rendering with single-pass architecture
  • Translator registry for customizing tag rendering
  • Output tuning options (tag ignoring, text replacements, link styles)
  • YAML front matter generation with metadata
  • Content hashing for deduplication
  • Heading-aware chunking for LLM processing
  • Whitespace normalization and cleanup
  • Telemetry hooks for performance monitoring
  • NDJSON transform for streaming pipelines
  • CLI tool for command-line conversion
  • TypeScript support with full type definitions
  • Dual ESM/CJS package exports
  • Comprehensive test suite
  • Performance benchmarks showing 3.75x faster than Turndown

Performance

  • Average processing time: 1.517ms (without Readability)
  • Average processing time: 7.731ms (with Readability)
  • Linear O(n) scaling confirmed across file sizes