Files
vslinx-ComfyUI-vslinx-nodes/web/docs/vsLinx_AppendLorasFromNodeToString.md
david 29102987e1 feat(LoRA to prompt): Node to get loras from power lora loader into img saver (#4)
* feat(LoRA to prompt): Node to get loras from power lora loader into prompt for image saver
2025-10-24 21:40:38 +02:00

26 lines
2.1 KiB
Markdown
Raw Permalink Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
Reads **LoRA entries from a Power LoRA Loader node** (by id/title or via the `powerloraloader_model` link) and appends them to your base prompt as **LoRA tokens** for metadata persistence with image-saver nodes. Tokens are formatted as `<lora:FILEPATH/FILENAME:STRENGTH>`, with **STRENGTH truncated to two decimals**.
This node does the following:
- Locates the target Power LoRA Loader by **node id**, **node title**, or by tracing the `powerloraloader_model` link from this node.
- Parses the target node’s **inputs** (not the MODEL object) and finds every `lora_*` entry (`{'on': bool, 'lora': str, 'strength': float}`).
- Builds a token for each entry in the exact format `<lora:PATH:STRENGTH>`, where `STRENGTH` is truncated (not rounded) to two decimals.
- Appends all tokens to the end of your input `text`, space-separated, preserving discovery order and removing duplicates.
Parameters:
| Parameter | Type | Description |
| -------- | ---- | ----------- |
| powerloraloader_model | MODEL (optional) | Connect the loader’s **model** output here so the node can trace the upstream link. |
| text | STRING | Base prompt text to which the LoRA tokens will be appended. |
| id | INT (optional) | Target Power LoRA Loader **node id**. If provided, this takes priority over `node_title`. |
| node_title | STRING (optional) | Manually edited **title** of the loader node. Used when `id` is 0 or no model is connected. |
| only_enabled | BOOLEAN | If true, include only LoRAs with `on: True`. Otherwise, include all `lora_*` entries that contain a valid path. |
| debug | BOOLEAN | Prints detailed logs (target resolution, discovered `lora_*`, built tokens, final output). Helpful for troubleshooting. |
Outputs:
| Parameter | Type | Description |
| -------- | ---- | ----------- |
| text | STRING | The input `text` with appended LoRA tokens like `<lora:Illustrious\Style\vslinxtybwbleach.safetensors:1.00>`. |
Notes:
- **Resolution priority:** `powerloraloader_model` → `id` → `node_title` link trace. If none resolve, the node returns the original `text` unchanged.
- **De-duplication:** Identical tokens are removed while preserving first-seen order.