Files
ComfyUI-Lora-Manager/py/recipes/constants.py
Will Miao c9f902a8af Refactor recipe metadata parser package for ComfyUI-Lora-Manager
- Implemented the base class `RecipeMetadataParser` for parsing recipe metadata from user comments.
- Created a factory class `RecipeParserFactory` to instantiate appropriate parser based on user comment content.
- Developed multiple parser classes: `ComfyMetadataParser`, `AutomaticMetadataParser`, `MetaFormatParser`, and `RecipeFormatParser` to handle different metadata formats.
- Introduced constants for generation parameters and valid LoRA types.
- Enhanced error handling and logging throughout the parsing process.
- Added functionality to populate LoRA and checkpoint information from Civitai API responses.
- Structured the output of parsed metadata to include prompts, LoRAs, generation parameters, and model information.
2025-05-06 21:11:25 +08:00

17 lines
285 B
Python

"""Constants used across recipe parsers."""
# Constants for generation parameters
GEN_PARAM_KEYS = [
'prompt',
'negative_prompt',
'steps',
'sampler',
'cfg_scale',
'seed',
'size',
'clip_skip',
]
# Valid Lora types
VALID_LORA_TYPES = ['lora', 'locon']