Skip to content

CHANGELOG.md

Latest commit

 

History

History
491 lines (283 loc) · 18.3 KB

CHANGELOG.md

File metadata and controls

491 lines (283 loc) · 18.3 KB

@lexbuild/core

1.25.0

Minor Changes

  • 6298069: Add single-pass multi-granularity conversion. The convert-usc and convert-ecfr commands now accept --granularities section,chapter,title[,part] together with --output-chapter, --output-title, and --output-part (eCFR only) to produce every requested granularity from one parse of the source XML. The builders' emitAt option accepts a ReadonlySet<LevelType> in addition to a single LevelType, and deeper levels emit first so that higher-level emissions see the complete subtree. Update scripts (update-usc.sh, update-ecfr.sh) collapse their N convert-* invocations into one, yielding a ~40-50% reduction in the incremental convert step's wall-clock on multi-granularity runs.

    Back-compat: the existing -g/--granularity + -o/--output single-granularity form continues to work unchanged and is mutually exclusive with --granularities.

1.24.1

Patch Changes

  • 28c1862: Fix eCFR title- and part-granularity emission. convert-ecfr -g title and convert-ecfr -g part previously emitted stub files containing only frontmatter and an empty heading, because the builder's emit condition used >= instead of strict equality on the level index — every intermediate level (section, part, chapter) emitted standalone instead of aggregating into the parent's children tree.

    Align EcfrASTBuilder with the USLM builder's strict-equality behavior and populate title_name from node.heading when the emitted node itself is a title (a title has no title ancestor). Title-granularity output grows from ~289 bytes per title to ~400 KB per title with full chapter/part/section content inlined; part-granularity output now includes section content.

1.23.3

Patch Changes

  • f1cf97f: Persist API aggregate snapshots during ingest so the Data API can serve stats and Federal Register hierarchy endpoints without recomputing full-table aggregates on live requests.

1.22.0

Minor Changes

  • Add @lexbuild/mcp package — Model Context Protocol server for LexBuild

1.19.0

Minor Changes

  • fe043bf: Add incremental content update pipeline and documentation

1.18.1

Patch Changes

  • bee35c5: Docs update; Version bump

1.18.0

Minor Changes

  • 434ecfe: Bump version for docs site version sync

1.17.3

Patch Changes

  • 66af7d2: Update documentation for Data API: README.md, docs/, and internal docs

1.17.2

Patch Changes

  • 47a48df: Use PBKDF2 for API key hashing and fix CI test runner
  • 0fef14e: Phase 5 of the Data API implementation. Adds full-text search across all three sources with faceted filtering, highlighting, and sort.
  • e63fd69: Refactor shared API key schema and hashing to @lexbuild/core
  • 92698e7: Phase 4 of the Data API implementation. Adds paginated document listings
  • 62e6ad0: Data API Phase 0 implementation
  • a94fa11: Phase 7 of the Data API implementation. Adds deploy script modes for API code and database deployment, and updates documentation across the monorepo.
  • a63e95e: Data API Phase 1
  • 2033419: Phase 3 of the Data API implementation.
  • 9b7b181: Phase 6 of the Data API implementation. Adds API key management with SQLite-backed storage, tiered rate limiting, usage tracking, and CLI commands for key lifecycle management.

1.17.1

Patch Changes

  • 16aae5b: Prettier Formatting

1.17.0

Minor Changes

  • 15c21c9: Meilisearch Docker Container

1.16.1

Patch Changes

  • ae8e01c: Update documentation and CLAUDE.md files

1.16.0

Minor Changes

  • deb4785: Enrich FR with API JSON metadata

1.15.3

Patch Changes

  • 62f5fd3: Prettier Formatting
  • 4599ccb: Cleanup and enhance comments throughout monorepo for better consistency

1.15.2

Patch Changes

  • c59e3a7: Patch Bump

1.15.1

Patch Changes

  • 428aaeb: Package README.md updates

1.15.0

Minor Changes

  • 0d01106: Implement Federal Register
  • 974392c: Implement bulk xml downloader for federal register

Patch Changes

  • c873f5c: Update packages CLAUDE.md files
  • 889a9c1: Fix parser error
  • 89a630b: Fix FR conversion
  • c3bec41: Implement concurrent downloads for FR
  • c612cb6: Fix formatting issues with FR raw and rendered text
  • 424d10c: Fix Federal Register downloader progress status; change 100ms request delay to 25ms for 4x faster throughput

1.14.1

Patch Changes

  • da51df3: bump version

1.14.0

Minor Changes

  • 9af2c94: @lexbuild/fr
  • f2118bd: Phase 1 implementing Federal Register source
  • 2a6e3f2: Implement Federal Register source

Patch Changes

  • d2e2520: Fix Claude PR Review Issues
  • 9a8a655: Update CLAUDE.md files
  • 67049b9: Update documentation for Federal Register source
  • c8e6e86: Implement fixes for Claude PR Review issues
  • 0c52c73: Fix ESLint errors

1.13.3

Patch Changes

  • a8b9cc1: Fix dependabot security vulnerabilities

1.13.2

Patch Changes

  • cab1c2c: Add download progress indicators
  • f6d7281: fix: address PR review findings for download progress

1.13.1

Patch Changes

  • f89a4ee: chore(deps): bump yaml from ^2.7.0 to ^2.8.3

1.13.0

Minor Changes

  • cb27ce5: Astro SEO enhancements, JSON-LD, Twitter card support, and robots.txt

1.12.0

Minor Changes

  • 17cf9bf: add list-release-points to CLI command and release point history API

1.11.0

Minor Changes

  • d22de4d: Bump version

1.10.1

Patch Changes

  • 59c12a1: feat: per title data resolution and retry logic for eCFR API downloads
  • 275f041: Address PR review feedback for download, highlights, and packaging
  • c735a6f: Update convert command summary footer for titles, chapters, parts, and sections
  • 82fb6cc: Manage Shiki's memory with forked child processes instead of a single process. Each child is 10k files and its memory is released back to the OS once highlighting is complete.
  • 7fde270: Implement hierarchical filenames for downloaded .md files (i.e. usc-title-01-chapter-01-section-1.md)
  • 95e44f5: eCFR data api fallbacks for when update is in progress; display date info during download
  • 0305588: Prettier formatting issue fixes; pnpm format

1.10.0

Minor Changes

  • 268dc81: Update eCFR downloader to handle govinfo eCFR XML or eCFR.gov API XML

Patch Changes

  • 2abf796: Update monorepo README.md
  • af2ea01: Fix eCFR frontmatter panel info
  • 791ca5e: Set the eCFR data API as the default eCFR source (govinfo still available if desired)
  • 65539ac: Update README.md files
  • d71a5d8: Auto detect latest release of the USC and add options to choose previous release points

1.9.4

Patch Changes

  • 6a7111a: Add ENFILE/EMFILE retry wrapper for file writes in converters

1.9.3

Patch Changes

  • 93b4e78: refactor: remove Next.js web application, add Astro SSR app for browsing converted legal content

1.9.2

Patch Changes

1.9.1

1.9.0

Minor Changes

Patch Changes

1.8.0

Minor Changes

1.7.0

Minor Changes

1.6.0

Minor Changes

  • #33 69b95ea Thanks @chris-c-thomas! - Change GitHub Actions to drop Node 20, keep 22, and add 24; Update documentation files to reflect those changes.

1.5.1

Patch Changes

  • 874a2de Thanks @chris-c-thomas! - fix: resolve workspace:* dependencies to actual versions in published packages

1.5.0

Minor Changes

1.4.2

Patch Changes

1.4.1

Patch Changes

1.4.0

Minor Changes

1.3.0

Minor Changes

  • #23 6ed94e2 Thanks @chris-c-thomas! - Fix incorrect output structure for chapter conversions, update web app, and modify convert command's summary stats.

1.2.0

Minor Changes

1.1.1

Patch Changes

1.1.0

Minor Changes

  • #15 b60c8a5 Thanks @chris-c-thomas! - Improve CLI help output with usage examples, granularity descriptions, and documentation links

  • #14 0968b27 Thanks @chris-c-thomas! - Add "title" output granularity (-g title) that produces a single Markdown file per title with recursive heading

Patch Changes

1.0.6

Patch Changes

1.0.5

Patch Changes

  • f86d74e: Quick fix

1.0.4

Patch Changes

  • 2598707: Fix README.md badge

1.0.3

Patch Changes

  • baaeba1: Update READMEs and package.json files

1.0.2

Patch Changes

  • 54219a7: Fix links that display on the npm registry

1.0.1

Patch Changes

  • Add package README.md files

1.0.0

Major Changes

  • First stable release. Converts all 54 titles of the U.S. Code from USLM XML to structured Markdown optimized for AI, RAG pipelines, and semantic search.

0.8.0

Minor Changes

  • Cleanup repo

0.7.0

Minor Changes

  • lexbuild convert --all now scans --input-dir for whatever usc{NN}.xml

0.6.2

Patch Changes

  • Cleanup files in project

0.6.1

Patch Changes

  • Enhance downloader for when all titles are downloaded

0.6.0

Minor Changes

  • 529f4bf: CLI TUI formatting fixes

0.5.0

Minor Changes

  • 3a29a8e: Add --titles multi-select option to download and convert commands. Supports ranges (1-5), comma-separated lists

0.4.1

Patch Changes

  • Add chalk, ora, and cli-table3 for polished terminal output with spinners and formatted

0.4.0

Minor Changes

  • Phase 4: Polish and Publish — snapshot tests, title-level README generation, CI/CD, documentation polish

0.3.0

Minor Changes

  • Phase 3: Scale & Download
    • feat(usc): OLRC downloader with zip extraction and release point support
    • feat(cli): lexbuild download command with --title, --all, --release-point
    • feat(usc,cli): --dry-run mode for convert command
    • feat(usc,cli): peak memory and token reporting in convert output
    • feat(core,usc): handle appendix titles with separate output directories
    • feat(usc): disambiguate duplicate section numbers with -2 suffix
    • feat(usc): status edge cases (repealed, reserved, transferred) in frontmatter and _meta.json

0.2.0

Minor Changes

  • abf4e13: Phase 2: Content Fidelity
    • Fix extra blank lines in multi-paragraph content blocks
    • Cross-reference link resolver with register/resolve/fallback and two-pass wiring
    • XHTML table conversion to Markdown pipe tables
    • USLM layout table conversion for TOC structures and tabular notes
    • Notes filtering with --no-include-notes, --include-editorial-notes, --include-statutory-notes, --include-amendments
    • _meta.json sidecar index generation at title and chapter levels
    • Chapter-level granularity mode (--granularity chapter)
    • Fix collector zone ordering bug (table/layout/toc checked before normal handlers)
    • E2E verified against Title 1 (39 sections) and Title 5 (1162 sections)

0.1.0

Minor Changes

  • Phase 1: Foundation — initial implementation
    • SAX streaming XML parser with namespace normalization
    • AST node types and stack-based builder with section-emit pattern
    • Markdown renderer with bold inline numbering, cross-reference link modes, notes, and blockquotes
    • YAML frontmatter generator with format versioning
    • USC converter pipeline: XML → parse → build → render → write section files
    • CLI convert command with output directory, link style, and source credit options
    • E2E verified against Title 1 (39 sections, 3 chapters)