refactor(core): restructure architecture for better maintainability

- Consolidate interface definitions into common.go for unified structure
- Merge OCR engine implementations into single engines.go file
- Consolidate utility functions from multiple files into filemanager.go and fileutils.go
- Streamline configuration management with environment variable overrides
- Enhance cross-platform compatibility and path handling
- Improve validation for input/output paths and file operations
- Implement two-tier logging system with critical vs detailed progress tracking
- Change CLI parameter from --llm_template to --llm-template for consistency
- Update documentation and dependency management

BREAKING CHANGE: Command-line parameter --llm_template renamed to --llm-template

Author: nodewee

CHANGELOG.md

@@ -0,0 +1,33 @@
+# Changelog
+
+All notable changes to this project will be documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [0.4.0]
+
+### Changed
+- Changed command-line parameter `--llm_template` to `--llm-template` for consistency with naming conventions
+
+### Improved
+- **Major Architecture Refactor**: Complete restructuring of core components for better maintainability
+- Enhanced cross-platform compatibility and path handling
+- Streamlined configuration management with environment variable overrides
+- Improved validation for input/output paths and file operations
+- Improved progress tracking with two-tier logging system (critical vs detailed)
+
+## [0.3.0]
+
+### Added
+- Core document text extraction functionality
+- Support for multiple file formats (PDF, images, e-books, office documents, HTML, text files)
+- OCR capabilities with Surya OCR and LLM Caller integration
+- Content-type strategy selection (text-first vs image-first processing)
+- Interactive tool selection with auto-detection
+- Cross-platform compatibility (Windows, macOS, Linux)
+- Resume capability for interrupted large document processing
+- Fallback extraction chains for robust processing
+- Comprehensive error handling and retry mechanisms
+- Structured logging with progress indicators
+- Build system with version injection and multi-platform binaries