[REFACTOR] Break down CodebaseToText god class into focused components

[jonasyr]

Description

The CodebaseToText class has grown to 600+ lines and handles multiple responsibilities (file traversal, pattern matching, content processing, output generation). This violates the Single Responsibility Principle and makes the code hard to maintain and test.

Current Issues

• Single class with 25+ methods
• Mixed concerns: file I/O, pattern matching, formatting, CLI handling
• Difficult to unit test individual components
• Hard to extend with new output formats or processing modes

Proposed Architecture

CodebaseToText (orchestrator)
├── FileTreeGenerator (directory traversal and tree generation)
├── PatternMatcher (exclusion pattern logic)
├── ContentProcessor (file reading and processing)
├── OutputFormatter (text/docx generation)
└── GitHubHandler (repository cloning and cleanup)

Implementation Plan

1. Extract PatternMatcher class first (lowest risk)
2. Extract OutputFormatter for text/docx generation
3. Extract FileTreeGenerator for directory operations
4. Extract GitHubHandler for Git operations
5. Refactor main class to use composition

Files Affected

• codebase_to_text/codebase_to_text.py (entire file)
• New files to create:
  • pattern_matcher.py
  • output_formatter.py
  • file_tree_generator.py
  • github_handler.py

Acceptance Criteria

• Each new class has a single, clear responsibility
• All existing functionality is preserved
• Code coverage remains ≥90%
• API compatibility is maintained for public methods
• Each component is individually testable
• Documentation updated for new architecture

Benefits

• Easier to add new output formats (PDF, HTML, etc.)
• Simpler unit testing of individual components
• Better separation of concerns
• Easier to modify pattern matching logic
• Cleaner code organization