Sd-parsers

- Improved modularity (Extractor and Parser modules can now be added and removed easily)
- Add support for parsing video metadata using ComfyUI default and WanVideoWrapper nodes.
- Add experimental support for png stealth metadata using the stenographic alpha channel extractor.
(thanks to nub-nub for the initial code)

**Full Changelog**: https://github.com/d3x-at/sd-parsers/compare/v0.5...v0.6

ComfyUI parser rework.
Improved metadata readability.
PromptInfo is now convertible to JSON.

**Full Changelog**: https://github.com/d3x-at/sd-parsers/compare/v0.4...v0.5

Initial support for images generated with [Fooocus](https://github.com/lllyasviel/Fooocus).
Usability improvements.

**Full Changelog**: https://github.com/d3x-at/sd-parsers/compare/v0.3.1...v0.4

Initial support for images generated with [Fooocus](https://github.com/lllyasviel/Fooocus).
Usability improvements.

**Full Changelog**: https://github.com/d3x-at/sd-parsers/compare/v0.3.1...v0.4rc1

* Speed improvements
* Added two_pass option to ParserManager to control parsing behaviour.
* Two-pass parsing tries to use the `info` property of the PIL image before falling back to `text`. This will considerable speed up parsing a large number of well-formed PNG files with actual SD parameters.
* If the majority of files is expected to not have SD metadata, setting two_pass to `False` will provide better performance.
* General code cleanup

> Read below and check the provided examples to get an idea of what has changed.