ComfyUI-NoelTextUtil

English | 한국어

Advanced text utility nodes for ComfyUI with LoRA trigger injection and unified prefix management.

[!NOTE] This project was created with a cookiecutter template. It helps you start writing custom nodes without worrying about the Python setup.

🙏 Special Thanks

NyaamZ - For the excellent Efficiency Nodes 💬ExtendeD project, which provided inspiration and reference patterns for the dynamic UI functionality and LoRA integration features in this project.

🚀 Quickstart

1. Install ComfyUI.
2. Install ComfyUI-Manager.
3. Look up this extension in ComfyUI-Manager. If installing manually, clone this repository under ComfyUI/custom_nodes.
4. Restart ComfyUI.

✨ Features

🎯 NoelUnifiedPrefix Node

• Character-based Prefix Management: Generates organized file prefixes for character portraits and animations
• Automatic Counter Management: Maintains separate counters for portraits, runs, and frames per character
• Smart Video Detection: Detects new animation runs based on time gaps
• Flexible Padding: Configurable zero-padding for consistent file naming
• State Persistence: Saves character counters between sessions

🔥 NoelLoRATriggerInjector Node

• LORA_STACK Integration: Works with LoRA_Stacker_ED to detect active LoRAs
• Dynamic Trigger Injection: Injects LoRA triggers into positive prompts based on active LoRAs
• Flexible Configuration:
  • Configurable LoRA slots (1-50, default: 5)
  • Append/Prepend injection modes
  • Deduplication support
  • Real-time LoRA slot management
• Enhanced UI/UX:
  • Dynamic node height adjustment
  • LoRA selection with search functionality
  • Real-time widget visibility management

🏗️ Architecture

Node Structure

NoelUnifiedPrefix

Inputs:
├── character_name: Character name for prefix generation
├── pad_portrait: Zero-padding for portrait counter
├── pad_run: Zero-padding for animation run counter
├── pad_frame: Zero-padding for frame counter
├── gap_timeout_sec: Time gap to detect new animation runs
├── portrait_image: Optional portrait image input
└── video_image: Optional video frame input

Outputs:
├── portrait_fullprefix: "{character}/portrait_{###}"
├── portrait_image: Pass-through portrait image
├── video_fullprefix: "{character}/animation_{###}"
├── video_image: Pass-through video image
├── frame_fullprefix: "{character}/animation_{###}/{#####}"
└── frame_image: Pass-through video image

NoelLoRATriggerInjector

Required Inputs:
├── lora_count: Number of LoRA slots to display (1-50)
├── dedupe: Remove duplicate triggers
└── mode: "append" or "prepend" trigger injection

Optional Inputs:
├── lora_stack: LORA_STACK from LoRA_Stacker_ED
├── positive_prompt: Base prompt for trigger injection
└── lora_X_name + lora_X_triggers: LoRA configuration slots (1-50)

Outputs:
├── lora_stack: Pass-through LORA_STACK
├── positive_prompt: Enhanced prompt with injected triggers
└── injected_triggers: List of injected trigger strings

JavaScript Extensions (lti.js)

• Dynamic Widget Management: Shows/hides LoRA slots based on lora_count
• Smart Height Adjustment: Automatically resizes node height using node.setSize()
• Enhanced LoRA Selection: Advanced dropdown with search and image previews
• External Extension Compatibility: Works with ed_betterCombos.js if available, falls back to basic functionality
• Real-time UI Updates: Immediate interface updates when lora_count changes

🔧 Usage Examples

NoelUnifiedPrefix - Character File Management

1. Portrait Generation: Connect portrait images to generate {character}/portrait_{###} prefixes
2. Animation Sequences: Connect video frames to generate {character}/animation_{###}/{#####} prefixes
3. Automatic Run Detection: New animation runs are detected based on gap_timeout_sec
4. State Persistence: Character counters are automatically saved and restored

NoelLoRATriggerInjector - LoRA Trigger Management

1. Connect LORA_STACK: Link from LoRA_Stacker_ED node
2. Set LoRA Slots: Configure lora_count and LoRA name/trigger pairs
3. Automatic Detection: Node automatically matches LORA_STACK with configured LoRAs
4. Trigger Injection: Triggers are injected into positive_prompt based on active LoRAs
5. Output: Enhanced prompt and LORA_STACK are passed to next nodes

Advanced Configuration

• Deduplication: Enable to prevent duplicate triggers in the final prompt
• Multiple LoRAs: Configure up to 50 LoRA slots for complex workflows
• Dynamic Adjustment: Change lora_count on-the-fly to show/hide slots
• Injection Modes: Choose between append (end) or prepend (beginning) trigger placement

🎨 UI Features

Dynamic Interface

• Responsive Layout: Node height automatically adjusts based on visible LoRA slots
• Smart Widget Management: Only shows the number of LoRA slots specified by lora_count
• Real-time Updates: Changes to lora_count immediately update the interface
• Automatic Resizing: Uses ComfyUI's native setSize() method for stable height management

Enhanced LoRA Selection

• Search Functionality: Quick LoRA name search with real-time filtering
• Image Previews: Visual LoRA previews when ed_betterCombos.js is available
• Dropdown Interface: Clean, intuitive selection interface
• Fallback Support: Works with or without external extensions
• Folder Integration: Uses ComfyUI's folder_paths.get_filename_list("loras") for LoRA discovery

🔌 Dependencies

Required

• ComfyUI (latest version recommended)
• Python 3.8+

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request. For major changes, please open an issue first to discuss what you would like to change.

Development Guidelines

• Follow existing code style and patterns
• Add tests for new functionality
• Update documentation for any API changes
• Ensure compatibility with ComfyUI's extension system

📄 License

This project is licensed under the GPL-3.0 License - see the LICENSE file for details.

📞 Support

• Issues: Use GitHub Issues for bug reports and feature requests

---

Made with ❤️ for the ComfyUI community