Changelog

All notable changes to vCon WTF will be documented in this file.

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

[Unreleased]

Added

  • Comprehensive CI/CD pipeline with GitHub Actions

  • Automated testing across multiple Python versions (3.10, 3.11, 3.12)

  • Security scanning with Bandit and Safety

  • Code quality checks with enhanced pre-commit hooks

  • Development Makefile with common tasks

  • GitHub issue and PR templates

  • Comprehensive contributing guidelines

  • Automated PyPI publishing workflow

  • Sphinx documentation framework

  • Read the Docs integration

  • GitHub Pages deployment

Changed

  • Enhanced pre-commit configuration with additional hooks

  • Updated development dependencies with security tools

  • Improved code style guidelines and enforcement

  • Updated documentation structure and content

Security

  • Added Bandit security scanning

  • Added Safety dependency vulnerability checking

  • Enhanced pre-commit hooks for security validation

[0.1.0] - 2024-01-XX

Added

  • Initial release of vCon WTF

  • Support for 6 major transcription providers: * Whisper (OpenAI) * Deepgram * AssemblyAI * Rev.ai * Canary (NVIDIA via Hugging Face) * Parakeet (NVIDIA via Hugging Face)

  • Comprehensive WTF format validation

  • Cross-provider testing framework

  • Performance benchmarking suite

  • Quality comparison system

  • Command-line interface with Rich UI

  • Bidirectional conversion (provider ↔ WTF)

  • Comprehensive test suite (100+ tests)

  • Integration testing with real API providers

  • Hugging Face model integration

  • Quality metrics and confidence scoring

  • Speaker diarization support

  • Timing accuracy validation

  • Extensible provider architecture

Features

  • Core Library: Pydantic-based WTF models with comprehensive validation

  • Provider Support: 6 major transcription providers with full bidirectional conversion

  • CLI Tool: Rich command-line interface with progress bars and detailed output

  • Cross-Provider Testing: Consistency, performance, and quality comparison across providers

  • Integration Testing: Real API testing with actual transcription providers

  • Quality Assurance: Multi-dimensional quality metrics and validation

  • Extensibility: Easy-to-add provider architecture for future integrations

Technical Details

  • Python 3.12+ support with modern type hints

  • Pydantic v2 for robust data validation

  • Click + Rich for beautiful CLI experience

  • Comprehensive Testing: Unit, integration, and cross-provider tests

  • Code Coverage: 50%+ overall coverage with 90%+ for core models

  • Security: Bandit scanning and dependency vulnerability checking

  • Documentation: Comprehensive README and inline documentation

Provider-Specific Features

  • Whisper: Log probability conversion, punctuation detection, quality metrics

  • Deepgram: Channel support, alternatives, speaker diarization

  • AssemblyAI: Utterances, sentiment analysis, auto-chapters, IAB categories

  • Rev.ai: Monologue elements, speaker confidence, timing accuracy

  • Canary: Hugging Face integration, audio transcription, model loading

  • Parakeet: Hugging Face integration, transducer-based processing

Cross-Provider Capabilities

  • Consistency Testing: Statistical analysis across providers

  • Performance Benchmarking: Speed, memory, and resource usage comparison

  • Quality Comparison: Confidence, accuracy, and completeness metrics

  • Report Generation: JSON and human-readable analysis reports

  • CLI Integration: Easy-to-use cross-provider testing commands

[0.0.1] - 2024-01-XX

Added

  • Initial project setup

  • Basic project structure

  • Core WTF models

  • Basic validation framework

  • Initial provider architecture

Release Notes

Version 0.1.0

This is the first major release of the WTF Transcript Converter library. It provides comprehensive support for converting between various transcription provider formats and the standardized IETF World Transcription Format (WTF).

Key Highlights:

  • 6 major transcription providers supported

  • Comprehensive cross-provider testing framework

  • Production-ready CLI tool

  • Extensive test coverage

  • Real API integration testing

  • Hugging Face model support

Getting Started:

pip install vcon-wtf
vcon-wtf --help

Cross-Provider Testing:

vcon-wtf cross-provider all input.json --output-dir reports/

This release establishes a solid foundation for transcription format standardization and cross-provider interoperability.

Version 0.0.1

Initial project setup with basic structure and core models.

Migration Guide

Upgrading from 0.0.1 to 0.1.0

Breaking Changes:

  • Updated Pydantic models to v2 format

  • Changed import paths for some modules

  • Updated CLI command structure

Migration Steps:

  1. Update imports:

# Old
from wtf_transcript_converter.models import WTFDocument

# New
from wtf_transcript_converter.core.models import WTFDocument

  1. Update CLI usage:

# Old
vcon-wtf convert input.json --provider whisper

# New
vcon-wtf to-wtf input.json --provider whisper

  1. Update validation:

# Old
from wtf_transcript_converter.validator import validate

# New
from wtf_transcript_converter.core.validator import validate_wtf_document

New Features:

  • Cross-provider testing framework

  • Performance benchmarking

  • Quality comparison

  • Enhanced CLI with Rich UI

  • Comprehensive documentation

Deprecation Notices

No deprecations in current version.

Security Advisories

No security advisories in current version.

Known Issues

  • Some providers may have rate limits that affect cross-provider testing

  • Large audio files may require significant memory for processing

  • Hugging Face models require internet connection for first-time download

Future Roadmap

  • Additional provider integrations (Google Cloud, Amazon Transcribe, Azure Speech)

  • Real-time transcription support

  • Advanced quality metrics

  • Machine learning-based quality assessment

  • Web-based interface

  • API server implementation

  • Database integration

  • Cloud deployment options

Contributing

See the Contributing guide for information on how to contribute to this project.

Support

  • Documentation: Check the full documentation

  • GitHub Issues: Report bugs and request features

  • Discord Community: Join our Discord for support

  • Email Support: Contact us at vcon@ietf.org

License

This project is licensed under the MIT License - see the LICENSE file for details.