Redesigned architecture to support multiple LLM services in Co-op Translator

[skytin1004]

This redesign allows for easier integration of new language and vision models while maintaining clean separation of concerns.

Solved #56

Purpose

This redesign aims to make the codebase more extensible and maintainable by:

• Creating a flexible architecture that supports multiple LLM and Vision providers
• Separating provider-specific implementations from core business logic
• Establishing interfaces for adding new language and vision models
• Organizing utilities and configurations in a more logical structure

Description

Major changes include:

1. Configuration System Restructuring

   • Moved provider-specific configs to dedicated modules
   • Created abstract base classes for configuration management
   • Implemented provider discovery and initialization logic

2. Service Layer Reorganization

   • Separated LLM and Vision services into provider-specific modules
   • Created abstract base classes for each service type

3. Utility Reorganization

   • Moved common utilities to utils/common
   • Grouped LLM-specific utilities under utils/llm
   • Grouped Vision-specific utilities under utils/vision

Related Issue

N/A - Architecture improvement initiative

Does this introduce a breaking change?

• Yes
• No

Type of change

• Bugfix
• Feature
• Code style update (e.g., formatting, local variables)
• Refactoring (no functional or API changes)
• Documentation content changes
• Other

Checklist

Before submitting your pull request, please confirm the following:

• I have thoroughly tested my changes: All core functionality remains intact with the new architecture
• All existing tests pass: Verified that existing tests work with the new structure
• I have followed the Co-op Translators coding conventions: Maintained consistent naming and structure
• I have documented my changes: Updated import paths and added inline documentation where necessary

Additional context

This refactoring sets the foundation for:

• Easy integration of new LLM providers (e.g., Anthropic, Google)
• Easy integration of new vision services (e.g., Google Cloud Vision)
• Better separation of concerns for future maintenance
• Clearer extension points for contributors

Functional testing complete. OpenAI model confirmed to work correctly. Test code will be added later.

[skytin1004]

I have reviewed the changes and everything looks good.