v0.3.0 - Major Architecture Improvements

CodeDox v0.3.0 - Major Architecture Improvements

🚀 Major Features & Improvements

Unified Code Extraction System

• Complete refactor to a unified extraction architecture with factory pattern
• New ExtractedCodeBlock model with semantic ExtractedContext for better code understanding
• Consolidated HTML, Markdown, and RST extraction into specialized extractors with shared base class
• Improved extraction rules: multi-line blocks always extracted, single-line with 3+ significant words
• Better handling of unclosed markdown fence blocks and HTML button elements

Source-Scoped Duplicate Detection

• Breaking Change: Duplicate detection now scoped to individual sources instead of global
• Same code snippets can exist across different documentation sources
• New composite unique constraint (document_id, code_hash) replaces global code_hash constraint
• Automatic database migration system with migration 008_remove_code_hash_unique
• Shared utility function find_duplicate_snippet_in_source() for consistent duplicate checking

Enhanced Frontend Experience

• Refactored DocumentDetail page with custom useDocumentDetail hook
• Improved search UX with better focus management and search result counts
• Enhanced language filtering with count display in dropdown

🔧 Technical Improvements

Extraction Engine

• Removed legacy SimpleCodeBlock and html_code_extractor.py systems
• Factory pattern for creating format-specific extractors
• Better context-aware HTML extraction with heading detection
• Improved title extraction combining section headings with page titles
• Enhanced RST support with better literal block detection

Database & Migrations

• Automatic migration application on app startup
• New migration check system that applies pending migrations automatically
• Updated SQLAlchemy models for new constraint structure

Code Quality

• Removed redundant wrapper functions and adapter layers
• Cleaned up 34 files with 2,805 additions and 1,921 deletions in main refactor
• Migrated all 269+ tests to new extraction API
• Added comprehensive test coverage for new extraction system

🐛 Bug Fixes

• Fixed UploadProcessor missing heartbeat tracking methods
• Removed pygame-specific corruption handling from RST extractor
• Fixed retry failed pages to maintain original source association

📝 Developer Experience

• Better extraction rules with semantic understanding
• Improved error handling and edge case coverage
• Enhanced test suite with dedicated extractor tests
• More consistent API across extraction formats

🔄 Migration Notes

Automatic Migrations: The database schema will be automatically updated when starting the application. The migration adds source-scoped duplicate detection.

Breaking Changes:

• Code extraction now uses ExtractedCodeBlock model instead of dictionary-based approach
• Duplicate detection API changed from global to source-scoped
• Removed container_type field from extraction models

---

Full Changelog: v0.2.8...v0.3.0

This represents a significant architectural improvement focusing on code extraction quality, database consistency, and user experience enhancements.