buchwandler

Changelog

[0.2.3] - Unreleased

[0.2.2] - 2026-10-11

Added

  • Added SpokenChange.language for backend-reported per-replacement language

  • Added provider-neutral token hints and progress events to bulk preparation

Changed

  • Derive package versions from Git tags and use generated metadata for runtime fingerprints

Fixed

  • Require release tag and distribution versions to match before publication

Documentation

  • Document the root Python API, including lexical review and exact-slice chunking

  • Match documented Spokenform bounds to package metadata

Quality

  • Guard root exports and packaged typing with tests and CI

[0.2.1] - 2026-10-11

Added

  • Added compiler-facing run preparation with exact source/output maps and provider-neutral linguistic hints

[0.2.0] - 2026-10-07

Added

  • Added public preparation, residual QA, reproducibility fingerprint, and JSON serialization APIs

Changed

  • BREAKING: Changed ttsready to a source-neutral, in-memory preparation API for caller-owned text units

Removed

  • BREAKING: Removed the 0.1 SSMD and document workflows, CLI, caches, speaker tooling, and file-based state

Documentation

  • Documented the migration, ownership boundary, Python API, packaging, and QA guidance for the 0.2 library boundary

[0.1.2] - 2026-10-03

Added

  • Added canonical SSMD and SSMD book loading with cached analysis and reproducible preparation

  • Added SSMD book workspaces, chapter-aware output, utterplan preparation, and txt export

Changed

  • Changed residual sequence handling and improved narrow-terminal CLI help

  • Changed sequence fallback settings for preparation, analysis, and reproducibility locks

Fixed

  • Fixed CRLF and CR normalization before SSMD parsing

Documentation

  • Documented installation, CLI, API, review, reproducibility, and release workflows

[0.1.1] - 2026-10-03

Added

  • Added per-artifact control for spelling or preserving residual sequences

Changed

  • Improved CLI help readability on narrow terminals

[0.1.0] - 2026-10-02

Added

  • Added canonical SSMD 0.9 and .ssmdbook input support

  • Added reports, cached context, and explicit overrides for speech review

  • Added logical speaker review and materialization as SSMD voice annotations

  • Added locks, verification, and freeze output for reproducible preparation

  • Added preview and TXT export from the shared structured TTS plan

  • Added a public Python API for loading SSMD and preparing previews

  • Added optional lexical and speaker integrations as package extras

  • Added Python 3.14 to the Python 3.10 through 3.14 CI matrix

Documentation

  • Documented installation, quickstart, CLI, review, reproducibility, speakers, and API boundaries