d3x-at/sd-parsers

A Python library to read metadata from images created by Stable Diffusion.

sd-parsers's Issues

image cannot be parsed

Hi,
this image failed to be parsed, prompt_data = parser_manager.parse(filename) returns None
I got the image from civitai.com, PNG info from automatic1111 can parse it.

Thanks.

Update import statement in README or refactor code.

When using pip to install the package, it is installed as sdparsers and not sd_parsers. Importing the library now requires sdparsers instead of sd_parsers. Please either refactor code or fix the documentation in README.md.

AttributeError: 'list' object has no attribute 'strip'

i`m trying to read comfyui metadata, and have this error:

PS D:\rep\pnginfo> File "C:\Users\User\AppData\Local\Programs\Python\Python310\lib\site-packages\sdparsers\parsers\comfyui.py", line 139, in _get_prompt positive_id, positive_prompt = get_prompt( File "C:\Users\User\AppData\Local\Programs\Python\Python310\lib\site-packages\sdparsers\parsers\comfyui.py", line 131, in get_prompt parts = list(self._get_parts(image_data, source_id, text_keys)) File "C:\Users\User\AppData\Local\Programs\Python\Python310\lib\site-packages\sdparsers\parsers\comfyui.py", line 163, in _get_parts yield node['inputs'][text_key].strip() AttributeError: 'list' object has no attribute 'strip'

Saving SD info to images

Would it be an idea to have the ability to save SD info to images?
Could have it save by which parser you choose?

I've been considering adding that to Promptvision, but your idea of using this as a lib is much better.